CHARLESTON, SC--Pleural fluid pH may not be nearly as helpful as physicians have thought for predicting outcome in patients with malignant pleural effusions.

"About 10 or so years ago, some clinicians discovered that a low pH in the pleural fluid was statistically associated with shorter survival," said John E. Heffner, MD, in a recent interview with Pulmonary Reviews. "Based on evidence from several studies, they decided that a low pH could be used to identify patients who are unlikely to live for very long and, thus, should not undergo pleurodesis."

These investigators also found that a low pleural fluid pH was associated with pleurodesis failure in patients with malignant pleural effusions. Many physicians therefore hypothesized that a low pH must be a sign of extensive pleural malignancy, which would make a successful pleurodesis unlikely. Because of these findings, expert opinion and standard texts, articles, and board review courses have all recommended using pleural fluid pH to make decisions about whether pleurodesis should be performed, said Dr. Heffner, dean and professor of medicine at the Medical University of South Carolina in Charleston.

For good reason, physicians have sought an accurate predictor of outcome in patients with malignant pleural effusions. Roughly 10% to 40% of attempts at pleurodesis fail; thus, an accurate predictor could help reduce costs and morbidity associated with this procedure. However, three new studies,[1-3] two of which were conducted by Dr. Heffner and his research team, have challenged the conventional wisdom regarding pleural fluid pH.

LIMITED PREDICTIVE VALUE

To assess whether pleural fluid pH really was a predictor of survival, Dr. Heffner and colleagues performed a meta-analysis of nine relevant studies.[1] There are two ways of conducting a meta-analysis: researchers can simply extract the data from each study summary "or you can do what we did and pool the primary data from each study," Dr. Heffner said.

They obtained information on 417 patients with malignant pleural effusions (mean age, 61 years). Lung and breast cancers were the most common primary neoplasms, affecting 43% and 18% of patients, respectively. The primary neoplasm was unknown in 15% of patients, and the remaining 24% had a variety of other malignancies, including ovarian, oral cavity, and lymphatic cancers.

Overall median survival was four months. It was only 2.8 months, however, for patients with lung, kidney, gastrointestinal tract, or other high-risk cancers.

Pleural fluid pH was independently associated with survival and remained so even when the researchers controlled for high-risk cancers. "We found a correlation--the lower the pH, the shorter the survival," Dr. Heffner said. "But there wasn't enough precision in the data to predict survival for individual patients."

Indeed, a pleural fluid pH of 7.28 or below was only moderately accurate at predicting survival. It correctly identified only 54.4% of all patients--and 62.7% of high-risk patients--who died within three months.

The findings were even less favorable in a study[2] by Craig M. Burrows, MD, and colleagues from the University of California at San Diego. Their analysis of 85 consecutive patients (mean age, about 60 years) with recurrent malignant pleural effusions showed that neither pleural fluid pH, pleural fluid glucose, age, nor extent of pleural carcinomatosis could accurately predict survival.

The only significant predictor of survival was a patient's score on the Karnofsky Performance Scale (KPS). Median survival was 395 days for those with a KPS score of 70 or more, versus only 34 days for those with a KPS score of 30 or less. The authors therefore recommend that, when considering whether pleurodesis is indicated in patients with malignant pleural effusion, physicians use the KPS to assess overall functional status and life expectancy.

POSTPROCEDURE OUTCOME

In a separate meta-analysis,[3] Dr. Heffner and his research team evaluated the predictive accuracy of pleural fluid pH for identifying pleurodesis outcome in patients with malignant pleural effusions. For this analysis, they were able to retrieve primary data from 11 studies and one letter, which yielded information on 433 patients with malignant pleural effusions. These patients were similar to those in the other meta-analysis with regard to age and the incidence and types of tumors.

Once again, Dr. Heffner's team found a correlation between the success rate of pleurodesis and pleural fluid pH: As pH decreased, so did the success rate. But the predictive value of this information turned out to be minimal. Even a pH as low as 7.15 or less had a positive predictive value of only 45.7% for pleurodesis failure. "We found that the pH doesn't predict pleurodesis failure in individual patients accurately enough to be clinically useful," Dr. Heffner told PULMONARY REVIEWS.

LOOKING AT THE LIMITATIONS

Dr. Heffner acknowledges that although his team's analyses were basically sound, they had some limitations. "A meta-analysis is only as good as the data that go into it," he emphasized, "and we found that a lot of the studies [in the meta-analysis] had design problems."

Most importantly, the data in the meta-analysis of pleurodesis outcome may have been biased because the researchers were not blinded to the pleural fluid pH test results. Furthermore, in each of the original studies, only one physician had decided whether pleurodesis succeeded or failed.

In the meta-analysis of pleural fluid pH and survival, several studies did not include consecutive patients--a limitation that may actually have made pleural fluid pH appear more accurate than it was. In addition, complete data were not available for all patients, and none of the studies reported instrument-related variability in pH measurements.

The study by Burrows and colleagues also had several limitations, which they noted in their report, including a small sample size and the failure to adjust for any effect of previous or current cancer treatment. And because the study only included patients with advanced cancer, the findings may not apply to those with earlier-stage disease.

Despite their limitations, these three investigations provide strong evidence that pleural fluid pH does not accurately predict survival or pleurodesis outcome in patients with malignant pleural effusions. "These patients should therefore not be excluded from pleurodesis just because of a low pleural fluid pH," Dr. Heffner concluded.

--Timothy Begany

References
1. Heffner JE, Nietert PJ, Barbieri C. Pleural fluid pH as a predictor of survival for patients with malignant pleural effusions. Chest. 2000;117:79-86.
2. Burrows CM, Mathews WC, Colt HG. Predicting survival in patients with recurrent symptomatic malignant pleural effusions. An assessment of the prognostic values of physiologic, morphologic, and quality of life measures of extent of disease. Chest. 2000;117:73-78.
3. Heffner JE, Nietert PJ, Barbieri C. Pleural fluid pH as a predictor of pleurodesis failure. Analysis of primary data. Chest. 2000;117:87-95.

Return to table of contents