ACSH Phthalate Panel:
C. Everett Koop, MD, ScD, Chair;
Daland R. Juberg, PhD;
Elissa P. Benedek, MD;
Ronald W. Brecher, PhD, CChem, DABT;
Robert L. Brent, MD, PhD;
Morton Corn, PhD;
Vincent Covello, PhD;
Theron W. Downes, PhD; Shayne C. Gad, PhD, DABT;
Lois Swirsky Gold, PhD;
F. Peter Guengerich, PhD;
John Higginson, MD, FRCP;
W. Hans K °nemann, PhD, RIVM;
James C. Lamb IV, PhD, DABT;
Paul J. Lioy, PhD;
Kimberly M. Thompson, ScD
To the Editor, Medscape*
We thank Dr. Schettler for his review[1] of the Report of the American Council on Science and Health (ACSH) Phthalate Panel[2] and for the opportunity to reaffirm our conclusions and to address his criticisms. All panelists received a copy of Dr. Schettler's letter, and no one expressed a desire to change our original report. This letter constitutes the panel's response to Dr. Schettler's concerns about the Panel Report. Nearly a year after its publication, we stand by our conclusions:
"DEHP [di-(2-ethylhexyl) phthalate] in medical devices is not harmful to even highly exposed people, those who undergo certain medical procedures such as regular hemodialysis or extracorporeal membrane oxygenation (ECMO). Furthermore, the Panel concludes that DEHP imparts a variety of important physical characteristics that are critical to the function of medical devices, and eliminating DEHP in these products could cause harm to some individuals."
Regarding Dr. Schettler's specific criticisms, we believe that he has mischaracterized both the evidence and our report. First, he claims that we selectively reported the scientific evidence. The safety of phthalates has been studied for more than 30 years, and nearly 1000 publications on DEHP appear on MEDLINE. Consequently any published review of this literature must be selective and focus on the weight of the evidence, as originally stated in our report. In fact, the Panel reviewed many more studies than we chose to cite. Although the Panel Report does not specifically cite the studies by Arcadi , Gray, Poon , and Li and their colleagues,[3-6] this omission does not indicate that the Panel failed to recognize the effects of DEHP on the testis. Based on review of the literature pertaining to reproductive effects of DEHP, the Report stated, "The target organ for DEHP reproductive toxicity in the rat appears to be the testis, for which adverse effects are well-documented (1). The reproductive effects are both strain- and species-specific, and young animals, in which most of the studies have been conducted, appear to be more sensitive than older animals (32)." The Panel did not intend to cite every available study pertaining to DEHP, but rather to review and summarize its findings in an effective and efficient manner.
Dr. Schettler seems to discount a trial that the panel considered relevant: Kurata and colleagues[7] studied adolescent marmoset monkeys during sexual maturation, a sensitive period for putative DEHP effects. Dr. Schettler states that "marmosets are sexually mature by 13 months of age," although this is not consistent with current thinking. Marmosets are considered to be sexually mature at 18 months of age and are more appropriately considered "adolescent" (late puberty to early adulthood) at 12 to 15 months of age[8]. Although it could be argued that the study by Kurata and colleagues does not assess the immature primate testis, it does provide a basis for assessing testicular effects in a nonrodent species. This was a point of interest to the Panel, given the relative rarity of having available data on toxicity in nonhuman primates, which are more difficult to study than are rats. In addition, the dose levels used by Kurata and coworkers (which were not associated with reproductive toxicity) have clearly been associated with reproductive toxicity in adult rats,[9] which underscores the point that primates seem to be less sensitive to DEHP-induced testicular effects.[7]
A standard approach in toxicology for evaluating testicular effects is to weigh the organ and then perform histopathologic examination (usually with light microscopy); if effects are seen, electron microscopy is done. Kurata and colleagues' study used a thorough procedure and still found no effects from DEHP. Finally, these researchers studied other correlates (eg, clinical chemistry) that are frequently used in conjunction with histopathologic determination of the toxicologic significance of a finding, and no changes in testicular zinc level or levels of testosterone or estradiol in the blood were reported for DEHP-treated animals.
Dr. Schettler states that the Panel Report failed to recognize and discuss Scientific Committee on Toxicity, Ecotoxicity, and the Environment of the European Commission decision to use, on the basis of developmental effects, a dosage of 3.7 mg DEHP/kg/day as a no-effect level in rodents. In fact, the testicular effects (ie, mild Sertoli cell vacuolation at 37 mg/kg/day) reported by Poon and colleagues,[6] the study on which the SCTEE dosage is based, were subtle. As noted in the Report, the "reproductive effects [for DEHP] are both strain- and species-specific..." and "testicular effects have been most pronounced in rats and mice, whereas other species appear to be either more resistant (i.e, hamsters; 33), or unaffected (i.e., marmoset monkey; 34)." Dr. Schettler seems to be most critical of the Panel's apparent failure to adequately acknowledge and address what he considers to be "permanent testicular damage" caused by DEHP. In fact, an important distinction of Arcadi and coworkers'[3] study, which reported testicular effects in rats after oral exposure, is that it describes aberrant spermatogenesis and testicular cell (seminiferous epithelium) condition in detail. The investigators attribute these findings to DEHP administration; however, the changes (eg, disorganization of the seminiferous epithelium and presence of round spermatids) can be indicative of delayed maturation and are reversible. This has been demonstrated in DEHP administration in neonatal rats.[10] The evidence suggests that the observations of abnormal spermatogenesis and testicular effects made by Arcadi and colleagues, particularly in the low-dose group, may not be treatment related or adverse.
The Panel also considered, but did not cite, other studies that support our findings concerning testicular toxicity of DEHP. For example, Agarwal and colleagues[11] reported effects of DEHP on the gonadal pathophysiology, sperm morphology, and reproductive performance of male rats after dietary exposure, but only at high doses. These effects have a threshold that must be considered. More pertinent to Dr. Schettler's concern about the "particular sensitivity of the immature developing testis to the toxicity of DEHP," we note the work of Sjoberg and colleagues[12] in which age-dependent responses of the rat testes were evaluated after both oral and intravenous (IV) administration of DEHP. These investigators reported that after administration of high oral doses, testicular damage was observed in immature animals but not in mature animals. Plasma concentration and urinary excretion data suggested that the gastrointestinal absorption of mono-(2-ethylhexyl)phthalate (MEHP) was higher in young animals. However, young rats were not more susceptible than older rats after repeated IV infusions of DEHP, suggesting that "the age-related difference observed in testicular response after oral administration of DEHP may be due to pharmacokinetic rather than tissue sensitivity differences. It is concluded that in assessing risks of testicular injuries in children exposed to DEHP, additional studies are required using species in which testicular development is more similar to that of humans."[12] This is precisely why the Panel focused on studies involving species closer in lineage to humans[7] and why it considered interspecies pharmacokinetic differences during its deliberations.
The purpose of the study by Ward and coworkers[13] was to elucidate and explore the role of a specific toxicologic mechanism in mediating toxicity. The Panel evaluated and chose to comment on the study in light of its findings related to the toxicologic mechanism, peroxisome proliferation, by which liver effects appear to be mediated. Important considerations when evaluating the findings of Ward and colleagues for relevance to humans include the high dose levels used (12,000 ppm for up to 24 weeks), the species (mice, which appear to be more sensitive than humans), and route of exposure (oral). The Panel did not ignore the finding as Dr. Schettler suggests, but decided to focus on the particular utility of this study in delineating the toxicologic mechanism for DEHP. The sensitivity of the species coupled with the large oral doses used in this study, in contrast to the lack of appreciable peroxisomal proliferation and lack of DEHP toxicity in humans, limits the utility of these results in human risk assessment.
In reference to a study by Crocker and colleagues,[14] Dr. Schettler states that "...when reviewing the renal toxicity of DEHP, the ACSH Panel fails to cite a study in which chronic oral dosing of rats with DEHP... showed reduced kidney function and cystic changes in the kidneys when compared to controls." In fact, we did cite this study, stating that "renal cysts and decreased kidney function in rats treated with DEHP have been reported (4)..." However, we note that Crocker and coworkers used very few animals, did not disclose the strain of the rat studied, and treated some of the animals with a leachate from an artificial kidney unit but did not measure the DEHP levels in these animals.
Dr. Schettler is correct that we did not cite the older studies of Sjoberg and Shneider and their colleagues[15,16] that discussed infant exposure to DEHP associated with invasive life-saving medical procedures. A review of Sjoberg and colleagues' study reveals that only 1 of 6 infants had a posttransfusion MEHP blood plasma level of 15 mcgg/mL and that the other 5 infants had, on average, posttransfusion levels of less than 4 mcg/mL, a substantial difference by any measure. To assess a larger infant population with exposure to DEHP, we cited a more recent study by Karle and coworkers,[17] which concluded that:
"Although significant concentrations of DEHP leach from the nonheparin-bonded circuits over time, our in vivo studies showed that the DEHP plasma concentrations were less than the previously reported values and do not correlate with an observable short-term toxicity."
Dr. Schettler makes two comments on misattribution of findings. In response to the first, we have realized that we inadvertently dropped a clause containing a reference in 1 sentence of the Report. The resultant sentence was confusing; however, the inference does not change. Following is a more clear version of that sentence: "In this study, as well as in one using baboons (47), evidence suggests that either the solubilizing agent (Tween surfactant) in which DEHP was dissolved or other confounding variables such as microemboli formation were important in the manifestation of toxicity."
Regarding the second claim, in the following sentence from the Report, references 5 and 4 cite the studies by Dunnill and Crocker and their colleagues[14,18]-- not the study by Karle and coworkers[17] as suggested by Dr. Schettler. The sentence stated, "Cystic kidney disease in hemodialysis patients (5) and renal cysts and decreased kidney function in rats treated with DEHP have been reported, (4) although in humans an association between DEHP and kidney effects is speculative and confounded by the medical condition of dialysis patients."
In response to Dr. Schettler's comments about metabolism, the Panel did consider the studies in question. The data Dr. Schettler mentions on percentages of metabolites from Albro and colleagues[19] are for the urinary metabolites that were hydrolyzed during the experimental analysis, not free in urine. Thus, these are glucuronides, and 18% of that fraction was MEHP (as shown in Table 7 of Albro and colleagues' study). The criticism is, therefore, unfounded. Clearly, free MEHP had to be transiently formed but was rapidly glucuronidated. Table 8 of the study shows 0% conjugates in rat urine and 80% in human urine. Dr. Schettler's criticism about immature glucuronidation pathways at birth is also unfounded. As emphasized at a January 2000 International Life Sciences Institute workshop on the subject of age and development, assumptions cannot be made about any developmental differences or similarities in drug metabolizing enzymes between rodents and humans. Notably, if rodents do not have this particular glucuronidation pathway as adults, how can ontogeny results be extrapolated to humans?
The Panel gave due consideration to the scientific data pertaining to interspecies metabolic differences and pharmacokinetic handling of DEHP. In our report, we noted the work of Rhodes and colleagues[20] that evaluated pharmacokinetic differences in rats and marmoset monkeys. The authors report that "in the marmoset, the excretion profile of [14C]-DEHP following oral, IP, and IV administration and the lower tissue levels of radioactivity demonstrated a considerably reduced absorption in this species compared to the rat." This again emphasizes the importance of animal model selection when evaluating the relevance of toxicologic data for humans and supports our emphasis and reliance on models that are more similar to humans.
Having the benefit of several physicians on the Panel who have seen first hand the evolution of medicine brought about by the use of phthalates in medical devices, the Panel wanted to emphasize that products containing phthalates provide important life-saving benefits. Anyone who has ever had an IV injection of a life-saving fluid or blood product stored in a collapsible IV bag should appreciate that these bags do not break easily or require the injection of sterile air to force fluid out, as did the glass bottles that preceded them. Further, we can now find and remove tumors and prevent cancer in parts of the colon that can only be reached by flexible plastic devices. Thousands of children every year owe their lives to the use of flexible vinyl tubing in ECMO. Every physician knows that the concept "first do no harm" strikes a delicate balance between the risks and benefits of treatment. Current standards of care demand that doctors save lives with as few side effects as possible first, and then address the side effects, if necessary. As medicine continues to evolve, we should expect standards of care to change, as devices and medications of the future that are demonstrated to be safer and more effective replace those used today.
We are hopeful that additional scientific investigations will improve insight into the degree of risk associated with DEHP in medical devices. The current review of 8 phthalates, including one of DEHP by the National Toxicology Program (NTP), may be an important step in this process. At the time of this writing, the NTP Review Panel had not reached a consensus,[21] although its preliminary findings on diisononyl phthalate (DINP)seem to be similar to the conclusions in our original report.[21]
A scientific analysis was done by the International Agency for Research on Cancer (IARC) inFebruary 2000 that used a weight-of-evidence approach (specifically considering toxicologic mechanism). This group reached conclusions similar to ours on whether DEHP is a carcinogen. The IARC downgraded its classification of DEHP from group 2b to group 3 (not classifiable as carcinogenic in humans) stating:
"In making its overall evaluation of the possible carcinogenicity to humans of DEHP, the working group took into consideration that (a) DEHP produces liver tumours in rats and mice by a non-DNA-reactive mechanism involving peroxisome proliferation; (b) peroxisome proliferation and hepatocellular proliferation have been demonstrated under the conditions of the carcinogenicity studies of DEHP in mice and rats; and (c) peroxisome proliferation has not been documented in human hepatocyte cultures exposed to DEHP nor in the livers of exposed non-human primates. Therefore, the mechanism by which DEHP increases the incidence of hepatocellular tumours in rats and mice is not relevant to humans."[22]
Finally, we agree with Dr. Schettler that there should be no tolerance for selective, false, and misleading reviews of scientific literature, failure to acknowledge and correct mistakes, nondisclosure of financial interests, or bias in publications that may be used to inform public policy. However, Dr. Schettler's comments do not provide convincing evidence or a compelling scientific basis for modification of our Report.
References
The Panel on A Scientific Evaluation of Health Effects of Two Plasticizers Used in Medical Devices and Toys*
* Signed by all members of the Panel, except for Dr. Lundberg, who as the Editor of Medscape prepared a separate response, and one member who was unable to participate in the response for health reasons.