 |
|
Clinical Efficacy of Devices for Nonablative Photorejuvenation
David J. Leffell, MD
Arch Dermatol. 2002;138:1503-1508.
ABSTRACT
| |
Objective To evaluate the scientific evidence for the efficacy of devices for nonablative photorejuvenation of the skin.
Data Sources All studies published between January 1996 and June 2002 in the dermatology literature listed in the MEDLINE database. Search terms included nonablative and photoaging or skin aging.
Study Selection All studies that presented data on human clinical trials that were designed to test the efficacy of nonablative light sources in reducing fine lines and wrinkles due to photodamage.
Data Extraction Data presented in the studies were reviewed and evaluated from the perspective of study design and validity of conclusions.
Data Synthesis Eleven studies were reviewed. Five different light sources were used, each with varying parameters. Ten different clinical end point scales and methodologies were used. No 2 studies used the same clinical end point standards. The majority of the studies provided no statistical analysis of data.
Conclusions The studies reviewed failed to present consistent data on the efficacy of nonablative photorejuvenation to improve or eliminate rhytids. The field is early in development, and continued improvement and standardization of study design are needed to determine the efficacy of these interventions.
INTRODUCTION
ONE OF the major advances in the field of procedural dermatology in the past decade has been the development of a range of laser technologies to address cosmetic and medical conditions of the skin. Most notably, the use of the carbon dioxide laser for facial resurfacing established a new frontier for understanding and treating conditions related to photodamage of the skin. In 1996, Fitzpatrick et al1 published a landmark article on ultrapulse carbon dioxide laser resurfacing of the skin. It is generally accepted that photoaging is in part due to degradation of collagen by metalloproteinases induced by exposure to UV radiation.2 Similarly, the improvement in facial rhytids that occurs after cutaneous injury, whether from dermabrasion or carbon dioxide laser, correlates with elevated dermal collagen levels and a diminution in elastotic material in the dermis.3 As a result of substantial promotion of carbon dioxide laser resurfacing to dermatologists and allied specialists, facial resurfacing became an extremely popular procedure. However, the usefulness of this technique is limited by prolonged healing time, risk of complications such as pigmentary disorders, and discomfort, including the need for regional and even general anesthesia.4 As the potential complications have become better known, the popularity of the procedure evidently has diminished. However, the demand for methods to reverse the cardinal signs of photodamage continues unabated.
The impact of neocollagenesis in repairing photodamage was first clinically appreciated with the systematic evaluation of the effect of topical tretinoin.5 That research established the fundamental standards for evaluating the clinical effectiveness of procedures or treatments intended to reverse the clinical signs of photoaging. The scientific evaluation of "antiaging" interventions, whether surgical or medical, presents special challenges in objective measurement. Quantification of improvement in rhytids is especially difficult, so rigorous study design, including clearly defined parameters for intervention and clinical end points, is essential.
The notion that the clinical manifestations of photodamage can be repaired by the application of selective wavelengths of light energy is a logical extension of the principles of nonphotoselective ablation. On the basis of clinical observation, several investigators began to study the effect of a variety of wavelengths, including selectively filtered polychromatic light, on photodamaged skin.
This study reviews the investigations on nonablative photorejuvenation that have been published to date in the peer-reviewed dermatology literature and attempts to determine the degree to which the efficacy of nonablative photorejuvenation is supported by the evidence developed in clinical trials.
METHODS
Studies reporting the results of clinical trials of nonablative photorejuvenation were identified in the MEDLINE database using photoaging and skin aging combined with nonablative as search terms. The latter term was selected to be as broad as possible to ensure that all published studies would be included. The following elements of the studies were tabulated: device; number of subjects; treatment parameters, including energy level and frequency of treatment; clinical end points; method of evaluation; evaluation of changes in dermal collagen; statistical methods; and authors' conclusions (Table 16-16).
|
|
|
|
Summary of Clinical Trials on Nonablative Photorejuvenation
|
|
|
RESULTS
Eleven studies were included in the analysis, representing 5 different types of light sources.6-16 Treatment regimens ranged from a single procedure6 to 8 treatments at 2- to 3-week intervals.11 In 10 of the trials, the clinical end point consisted of observation only.6-15 A variety of different scales were used to rate the clinical results. Also, depending on the study, rhytids, telangiectasias, and/or "skin texture" was evaluated. In most cases, fine lines and rhytids represented the clinical end point. In one study, profilometry of silicone molds was performed.16 In 6 of the studies, blinded independent observers were used to evaluate clinical photographs before and after treatment.6-7,10, 12, 15-16 In one study, only the treating physician and the patient evaluated the clinical benefit,14 and in another study, only the subjects graded the effect of the treatment.13 In one study, the participating physicians performed the evaluations of clinical improvement.9 In 55% of the studies, no statistical analysis of data was conducted. Histologic analysis was the predominant method of evaluating changes in dermal collagen, with or without special stains. In one study, biochemical analysis was performed to determine the activity of collagen precursor molecules.7 Improvement in clinical end points was highly variable and ranged from no improvement10 to some improvement6-9,11-12,14-16 in patients with mild or moderate wrinkles. It should be noted that not all claims to clinical improvement were supported by statistical analysis, nor were the conclusions uniformly based on objective evaluation. In one case, there was no correlation between histologic improvement and clinical change.11
PULSED DYE LASER
In 1999, Zelickson et al6 reported on pulsed dye laser therapy for sun-damaged skin. They evaluated the effect of a single pulsed dye laser treatment at a 585-nm, 450-millisecond pulse on 20 patients. The energy density was 3 to 6.5 J/cm2. The objective of the study was to assess the effect of laser treatment on the clinical appearance of facial wrinkles and the histologic appearance of elastin and collagen. Half of the patients had mild to moderate sun-induced rhytids, and the other half had moderate to severe sun-induced rhytids. Photographs of patients before and after treatment were evaluated by 3 blinded observers. Rhytids were rated on a 5-point scale, and the scores were then averaged. Biopsy specimens were obtained to evaluate the effect of laser on collagen production. Nine of 10 patients with mild to moderate wrinkling showed an improvement of 50% or more, with 30% showing improvement of 75% or more. Maintenance of this effect was variable over the 12-month follow-up period. Histologic analysis revealed a thickened stratum spinosum and a thickened layer of normal-staining collagen within the superficial dermis. Downward compression of basophilic connective tissue was also seen. Ultrastructural evaluation showed an increased amount of normal-appearing elastic and collagen fibers within the superficial dermis. No statistical analysis of data was presented in this qualitative study.
Bjerring et al7 hypothesized that enhanced collagen production could result from laser targeting of the microvasculature. Two groups of volunteers, totaling 40 subjects, were involved in the study. One group of 10 subjects was selected for biochemical analysis of the laser-treated site. The second group of 30 subjects was treated with identical laser parameters for facial wrinkles, and the cosmetic outcome was evaluated. Seventy-two hours after treatment, the concentration of aminoterminal propeptide of type III procollagen was measured. Cosmetic efficacy was judged through an evaluation of photographs. Standardized reference photographs were used to orient the evaluators to the grading system, and clinical photographs were randomized and independently rated in a blinded fashion. There was an increase in procollagen production at the treatment site, but lower procollagen production at sites that were multiply treated. The lower increase in collagen production at the double-treated sites suggested that the additional energy dosage had a negative physiologic effect. The authors also concluded that there was a cosmetic improvement in subjects, with an average value of 1.88 reduction in wrinkles as measured according to the Fitzpatrick Wrinkle Severity Scale.
Rostan et al8 compared the effect of treatment with low-fluence long-pulse dye laser and coolant with that of coolant treatment alone. Patients were randomly assigned to receive full-cheek laser treatment on one side of the face and coolant alone on the other. Collagen was evaluated for type I procollagen staining in biopsy specimens obtained from 4 patients who were randomly chosen. The method of clinical evaluation is not defined: it is not possible to tell whether the ratings are self-evaluations by the patient, ratings by the treating physician, or evaluation by blinded observers studying preoperative and postoperative photographs. The authors cite a statistically significant improvement (P = .004) in rating from the pretreatment to the posttreatment period in the laser-treated areas, compared with the control side, but do not appear to compare the treated side with the control side in each patient in a paired fashion.
ND:YAG LASER
Menaker et al9 reported on a series of 12 patients treated with the Nd:YAG laser. Three months after treatment, 4 of 10 patients showed a 1-point improvement on a 6-point scale in periocular rhytid severity. Biopsy specimens were obtained from all patients and assessed for collagen production by special stains. At 1 month, there was a 1-point (on a 6-point scale) increase in collagen of the papillary and reticular dermis in 5 of 10 patients (P<.06). The changes did not persist at 3 months. Also, Menaker and colleagues reported hyperpigmentation (in 4 patients) and pitted scarring (in 3 patients) as adverse effects.
Kelly at al10 reported on the use of cryogenic spray cooling with the Nd:YAG laser (1320 nm) to provide epidermal protection, while causing selective photocoagulation of the upper dermis. Thirty-five adults with periorbital rhytids were treated 3 times at 2-week intervals. Rhytids were categorized into 1 of 3 groups that were determined by the mean pretreatment score of mild, moderate, or severe as evaluated by 3 dermatologists who are experienced in laser treatment but who were not involved in the study. A paired comparison was made from photographs of each site. Photographs in this study were standardized for exposure, frame of reference, and distance. Differences between the pretreatment and posttreatment mean scores were evaluated by a paired t test performed for each group. Final assessment at 24 weeks after the last treatment showed statistical improvement only in the group with severe rhytids.
Trelles et al11 studied the effect of the Nd:YAG at 1320 nm using a cryogenic delivery system and a skin temperature sensor. Ten patients were studied. Two weekly treatments were given over 4 weeks. Standardized clinical photography was performed. Efficacy was rated by improvement in the skin condition class ranging from complete removal of wrinkles to little or no improvement. Evaluation was performed by the treating clinician. Biopsy specimens were obtained from all patients before and after treatment. Posttreatment skin biopsy specimens, compared with pretreatment specimens, revealed thickening of the epidermis and an increase in the number and density of collagen fibers in the dermis. At 10 weeks after the first treatment session, only 2 patients were satisfied with the results. The remaining 8 patients expressed little or no satisfaction with the results, observing very little or no effect. However, the histologic findings in the 2 patients who expressed satisfaction were not judged better than those in the other 8 patients. There was no statistical analysis in this study.
Goldberg12 reported on full-face nonablative dermal remodeling with a 1320-nm Nd:YAG laser. Ten patients were treated 5 times at 3- to 4-week intervals. All subjects were evaluated by a nontreating independent observer for the degree of improvement 6 months after the last treatment. Improvement, based on examination of 35-mm photography, was divided into quartiles: no improvement, some improvement, substantial improvement, and total improvement. Biopsy specimens were obtained from 4 randomly chosen subjects before and after treatment. Biopsy sites were treated with the same parameters used for treated facial skin. Histologic examination by a board-certified dermatopathologist who was unaware of the study design evaluated the amount and homogenization of collagen. At 6 months after the final treatment, all 10 subjects reported at least some improvement in their skin. The evaluation of the nontreating physician showed no apparent improvement in 4 subjects, some improvement in 4, and substantial improvement in 2. No subjects showed total improvement, but all subjects showed histologic evidence of some degree of new collagen formation. No statistical analysis was provided. The photographs were illuminated in a variable fashion and presented at different distances. New collagen formation did not necessarily correlate with clinically obvious improvement.
INTENSE PULSED LIGHT
Bitter13 reported on noninvasive rejuvenation of sun-damaged skin using full-face intense pulsed light treatments. Forty-nine patients with varying degrees of sun damage were treated with a series of 4 or more full-face treatments at 3-week intervals. The average number of treatments was 4.9. Subject evaluation and skin biopsy findings were used to assess treatment results. Bitter reported that all aspects of photodamage, including fine wrinkling, skin smoothness, skin laxity, irregular pigmentation, redness, and flushing, showed visible improvement in more than 90% of subjects. Eighty-eight percent of patients were satisfied with the overall results. The results of the patient self-assessment are reported without statistical evaluation of significance. Clinical photographs in the article are difficult to assess as the postoperative photographs appear overexposed. Photomicrographs of skin biopsy specimens are also difficult to interpret because of the level of magnification. There is no discussion or analysis of the histologic findings other than a qualitative description. From Bitter's article, it is not possible reach a definitive conclusion regarding the actual effect of intense pulsed light on photodamaged skin.
Negishi et al14 reported on the use of intense pulsed light for the rejuvenation of skin in an Asian population. A total of 97 patients between the ages of 22 and 70 years were treated, and the results were rated by both patients and physicians according to a 4-point scale. The evaluation does not appear to have been blinded on the part of the physicians, nor is there any statistical evaluation or intergroup comparison. A single before-and-after photograph of a skin biopsy specimen is shown. The authors conclude that more than 65% of patients provided a rating of good or excellent for skin texture.
Goldberg and Cutler15 studied 30 patients treated with intense pulsed light and rated the clinical effect on a 4-point scale. Two independent nontreating observers rated photographs of subjects before and after treatment. No biopsies were performed, and no statistical analysis of ratings is provided, but the authors conclude that all subjects showed some improvement, while none showed total resolution of wrinkles.
ERBIUM:GLASS
In an article on the use of the erbium:glass laser (1540 nm), Fournier et al16 reported on nonablative remodeling in 60 patients. Patients were treated 4 times at 6-week intervals and evaluated using digital photographs. Photographs of all patients were taken in a systematic fashion before, immediately after, and 7 days and 6 weeks after the first treatment and throughout the subsequent treatments. Collagen production was evaluated with silicone molds and ultrasound measurement. Ultrasound revealed a 17% increase in dermal thickness (P<.05). Silicone imprints were analyzed and anisotropy was rated. There was a 40% improvement in dermal collagen as noted by the reduction in anisotropy (P<.001). Six weeks after the fourth treatment, 62% of patients were satisfied (score of 3 on a scale according to which 4 meant "very satisfied"). Clinical observer data are provided in a qualitative fashion.
COMMENT
Eleven studies were reviewed involving 5 different light sources. The variety of devices, range of study device parameters, and the absence of a standardized means of evaluating improvement in photodamaged skin make it difficult to draw conclusions about the efficacy of these devices for nonablative photorejuvenation. Even if the cumulative data on only 1 particular device are considered, wide variation in study design, including treatment frequency and timing, makes it difficult to conclude whether nonablative photorejuvenation is effective.
There are several questions that arise as data for nonablative photorejuvenation are assessed:
1. Does the technology (laser and/or intense pulsed light) have a demonstrably objective beneficial effect in the treatment of fine lines and rhytids related to photoaging?
The studies discussed herein begin to suggest a possible benefit for the devices studied, but no single study provided objectively evaluated, statistically validated data to support the notion that the effect is clinically significant and reproducible.
2. If such a claim is made, what are the clinical end points that are established in the study and how are those end points evaluated?
The most common end points were improvement in fine lines and wrinkles or skin texture. However, because no standardized approach to rating was used, it is not possible to compare the results of studies for particular devices. The use of nonblinded observers and treating physicians as evaluators undermines the scientific basis of the studies in which those approaches to rating were used.
3. How do these new treatments compare in efficacy (including risk of adverse effects and degree of improvement relative to number of treatments) to established medical and surgical treatments?
None of the studies compared nonablative photorejuvenation with other established methods of improving photodamaged skin. This is understandable given the very early stage of development of this photorejuvenation technology. However, a comparative study should be performed to confirm whether the results from any of these techniques are preferable to existing treatment options, such as tretinoin, chemical peel, or carbon dioxide laser ablation. Should the claims for nonablative photorejuvenation be substantiated, comparison with existing approaches will certainly be required to determine whether adoption as standard of care is warranted.
4. What are the ideal parameters for a given nonablative photorejuvenation device?
The studies discussed herein acknowledge the need to identify ideal treatment parameters. The large number of variables, including energy delivery parameters and patient treatment schedules, make it especially challenging to define optimal treatment conditions. Ideally, optimal treatment parameters will be defined by the science of neocollagenesis.
5. Is it necessary to prove that nonablative photorejuvenation has an objectively evaluable effect on the presence of fine lines and wrinkles as long as the patient feels it is beneficial?
This question goes to the crux of cosmetic dermatology. Beauty is in the eyes of the beholder, and clinical improvement is ultimately measured in the patient's mirror. However, an ethical issue arises because of the strong impact of "suggestion" on patient perception. It would be injudicious for dermatologists to suspend their scientific commitment to objective validation of clinical effects simply because the patient has come to believe that he or she sees an improvement.
In the studies reviewed herein, substantial effort was made to evaluate the improvement induced by the treatment. In the majority of cases, a rating scale of rhytids, sometimes including skin texture and telangiectasias, was used. Because the clinical end points in improvement of photodamage are so difficult to evaluate, special effort must be taken to provide as objective an evaluation as possible. One reasonable structure for evaluation could include at least 3 observers, trained in a similar fashion to rate photodamaged skin, who independently evaluate preoperative and postoperative photographs that are standardized in all photographic parameters. It was not possible in most cases to draw conclusions from the photographs presented in the articles that were reviewed, suggesting that even if clinical photographs are consistent and of the highest quality, it is essential that journal reproduction reflect that accuracy in order to permit a full evaluation of the claims.
For purpose of study comparison, it is necessary to standardize treatment regimens and energy levels or at least to have a study design that includes a reliable control structure. In the studies reviewed, the number of treatments varied from 1 to as many as 8. Given the theoretical consideration that it is the effect on small vessels that is leading to new collagen formation, it appears that there must be some ideal treatment regimen. In the case of pulsed dye treatment of vascular lesions, the end point is obvious and multiple treatments are typically required. Treatment ends when the vessel disappears or the patient is satisfied.
The scientific basis for nonablative photorejuvenation is presumably new collagen formation. In the majority of studies, only a few biopsy specimens were obtained, stained with hematoxylin-eosin or special stains, and subjectively evaluated for the presence of new collagen. The time course to biopsy varied. In only one study was biochemical analysis of a collagen precursor performed. Despite the importance of understanding the impact of light energy on collagen, it is interesting to note that in one study there was improvement in the histologic findings in all 10 cases, but only 2 of 10 patients expressed satisfaction with the results of treatment.11
Based on the data reviewed herein, it is not possible to conclude that there was a meaningful clinical improvement in treated patients, independent of the device that was used. One could infer a trend that suggests that patients receive some minor benefit, although it is not known if it is sustained or what the ideal parameters of treatment are. The most objectively designed study showed a statistical improvement of P<.05 only in the group with severe rhytids. Overall, it should be noted that the incidence of adverse effects was very low.
CONCLUSIONS
Nonablative photorejuvenation represents a new approach to the treatment of the damaging effects of UV radiation on skin. Previous laser approaches have resulted in a relatively high incidence of adverse effects. Nonablative photorejuvenation promises to stimulate neocollagenesis and to yield clinical improvement, absent the risks associated with more destructive approaches. While it is true that nonablative photorejuvenation is a new area of clinical exploration, the data in the studies published to date make it difficult to draw conclusions about its effectiveness. While the clinical end points are typically subject to imperfect assessment, dermatologic investigation has established research methods for measuring clinical photoaging. Investigators in this promising area should seek to emulate those scientific precedents. One group has reported on a digital system for evaluating the effect of nonablative photorejuvenation on skin, but only 2 patients were studied.17 The field of nonablative photorejuvenation is further hampered by the large number of devices that are being used for this purpose. Comparative studies, which are the "gold standard" for any procedure, are difficult to construct and require a large number of subjects to achieve statistically useful information. However, the variety of devices, range of treatment parameters, and complexity of clinical end point analysis should not be deterrents to designing scientifically sound studies that evaluate the effect of light sources on the cardinal signs of photoaging.
AUTHOR INFORMATION
Accepted for publication July 17, 2002.
Corresponding author and reprints: David J. Leffell, MD, Department of Dermatology, Yale University School of Medicine, PO Box 208059, New Haven, CT 06520-8059.
From the Department of Dermatology, Section of Dermatologic Surgery and Cutaneous Oncology, Yale University School of Medicine, New Haven, Conn.
REFERENCES
| |
1. Fitzpatrick RE, Goldman MP, Satur NM, Tope WD. Pulsed carbon dioxide laser resurfacing of photo-aged facial skin. Arch Dermatol. 1996;132:395-402.
FREE FULL TEXT
2. Fisher GJ, Wang ZQ, Datta SC, Varani J, Kang S, Voorhees JJ. Pathophysiology of premature skin aging induced by ultraviolet light. N Engl J Med. 1997;337:1419-1428.
FREE FULL TEXT
3. Stuzin JM, Maker TJ, Baker TM, Kligman AM. Histologic effects of the high-energy pulsed CO2 laser on photoaged facial skin. Plast Reconstr Surg. 1997;99:2036-2050.
ISI
| PUBMED
4. Fitzpatrick RE. Laser resurfacing of rhytides. Dermatol Clin. 1997;15:431-447.
FULL TEXT
|
ISI
| PUBMED
5. Gilchrest BA. Treatment of photodamage with topical tretinoin: an overview. J Am Acad Dermatol. 1997;36(suppl):s27-s36.
6. Zelickson BD, Kilmer SL, Bernstein E, et al. Pulsed dye laser therapy for sun damaged skin. Lasers Surg Med. 1999;25:229-236.
FULL TEXT
|
ISI
| PUBMED
7. Bjerring P, Clement M, Heickendorff L, Egevist H, Kiernan M. Selective non-ablative wrinkle reduction by laser. J Cutan Laser Ther. 2000;2:9-15.
FULL TEXT
| PUBMED
8. Rostan E, Bowes LE, Iyer S, Fitzpatrick RE. A double-blind, side-by-side comparison study of low fluence long pulse dye laser to coolant treatment for wrinkling of the cheeks. J Cosmet Laser Ther. 2001;3:129-136.
9. Menaker GM, Wrone DA, Williams RM, Moy RL. Treatment of facial rhytids with a nonablative laser: a clinical and histologic study. Dermatol Surg. 1999;25:440-444.
FULL TEXT
|
ISI
| PUBMED
10. Kelly KM, Nelson JS, Lask GP, Geronemus RG, Bernstein LJ. Cryogen spray cooling in combination with nonablative laser treatment of facial rhytides. Arch Dermatol. 1999;135:691-694.
FREE FULL TEXT
11. Trelles MA, Allones I, Luna R. Facial rejuvenation with a nonablative 1320 nm Nd:YAG laser: a preliminary clinical and histologic evaluation. Dermatol Surg. 2001;27:111-116.
FULL TEXT
|
ISI
| PUBMED
12. Goldberg DJ. Full-face nonablative dermal remodeling with a 1320 nm Nd:YAG laser. Dermatol Surg. 2000;26:915-918.
FULL TEXT
|
ISI
| PUBMED
13. Bitter PH. Noninvasive rejuvenation of photodamaged skin using serial, full-face intense pulsed light treatments. Dermatol Surg. 2000;26:835-842.
FULL TEXT
|
ISI
| PUBMED
14. Negishi K, Tezuka Y, Kushikata N, Wakamatsu S. Photorejuvenation for Asian skin by intense pulsed light. Dermatol Surg. 2001;27:627-631.
PUBMED
15. Goldberg DJ, Cutler KB. Nonablative treatment of rhytids with intense pulsed light. Lasers Surg Med. 2000;26:196-200.
FULL TEXT
|
ISI
| PUBMED
16. Fournier N, Dahan S, Barneon G, et al. Nonablative remodeling: clinical, histologic, ultrasound imaging, and profilometric evaluation of a 1540 nm Er:glass laser. Dermatol Surg. 2001;27:799-806.
FULL TEXT
|
ISI
| PUBMED
17. Friedman PM, Skover GR, Payonk G, Kauvar AN, Geronemus RG. 3D in-vivo optical skin imaging for topographical quantitative assessment of non-ablative laser technology. Dermatol Surg. 2002;28:199-204.
FULL TEXT
|
ISI
| PUBMED
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati
What's this?
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
Evaluation of Acne Scar Treatment With a 1450-nm Midinfrared Laser and 30% Trichloroacetic Acid Peels
Carniol et al.
Arch Facial Plast Surg 2005;7:251-255.
ABSTRACT
| FULL TEXT
Nonablative Laser and Light Therapies for Skin Rejuvenation
Kim and Geronemus
Arch Facial Plast Surg 2004;6:398-409.
ABSTRACT
| FULL TEXT
Single-Pass Carbon Dioxide Laser Skin Resurfacing Combined With Cold-Air Cooling: Efficacy and Patient Satisfaction of a Prospective Side-by-Side Study
Raulin and Grema
Arch Dermatol 2004;140:1333-1336.
ABSTRACT
| FULL TEXT
Energy Delivery Devices for Cutaneous Remodeling: Lasers, Lights, and Radio Waves
Alam et al.
Arch Dermatol 2003;139:1351-1360.
FULL TEXT
|
