Extracts from the Manual of Procedures for the 1995-8 LIGHT-ROP clinical trial
by Drs. Reynolds, Spencer, et al.
You are on page
1 2 3 4 5 6 7 8 9 10 11
Go straight to some highlights on this page:
The authors misquote a prior study
The authors criticise another study for late protection
Light reduction may protect against more severe ROP
A comparison of light levels and ROP gets suppressed
Authors allege that oxygen therapy is controllable
Light toxicity and oxygen toxicity are additive
> Page 2-18 > continued 06-07-95
2.6.4 PREVIOUS CLINICAL STUDIES
OF LIGHT AND ROP
Before the importance of the role of inspired oxygen levels was recognized in ROP, two studies addressed the question of whether or not light exposure is a factor in the development or severity of this disease. Hepner et al. patched the eyes of five premature > Page 2-19 > 06-07-95 infants from birth until they weighed 2000 grams.105 [Note: Hepner et al. did not patch from birth but only within 24 hours.] They found that four of the five infants developed ROP and they concluded that light was not a factor in the development of ROP.105 Locke and Reese patched one eye in each of 22 premature infants (birthweight less than 2000 grams).106 They found that there was no difference in the incidence of ROP between the patched eyes and the unpatched eyes.106
Although pioneering and innovative at that time, the studies by Hepner et al. and Locke and Reese are unsatisfactory by today's standard for clinical tnals. Neither study used large enough sample sizes. Neither series used randomization of treatment assignment. The treatments were not masked from the observer and no standard classification of ROP existed at the time. Also, early examination during the "active" phase of ROP was not as frequently employed as it is now. Perhaps, most important, both studies were performed before the role of oxygen in the production of ROP was appreciated. Contrary to current practices, the use of high levels of inspired oxygen in the 1940s and early 1950s may have overshadowed any potentially beneficial effects of light reduction produced by these studies.
In 1985, a study by Glass et al. suggested that the light in the neonatal nursery may have some role in either the development or severity of ROP.107 This study concluded that "a high level of ambient illumination commonly found in the hospital nursery may be one factor contributing to retinopathy of prematurity." The authors based this statement on their work comparing the incidence of ROP in infants who were exposed to standard ambient nursery light to the incidence of ROP in infants exposed to reduced light levels. They found that the incidence of ROP was lower in those infants who were "protected" from the high levels of ambient light in the nursery.
There are some deficiencies in Dr. Glass' study.108 The infant subjects of the study were not randomized. The light exposure was not controlled nor accurately measured and the treatment was not masked. The ophthalmoscopic monitoring protocol was not described in detail. The evaluation of the occurrence of ROP was not carried out via a protocol to assure masking. The study107 utilized a neutral density filter but provided no measurement of its transmission characteristics except that it "reduced the intensity of > Page 2-20 > 06 07 95 light at the infant's face by approximately 50 percent." Dr. Glass seems to have recognized these limitations and concludes the publication:
". . . this study suggests that the levels of light common in the hospital nursery may contribute to the incidence of oxygen-induced retinopathy of prematurity, especially in infants weighing less than 1000 grams at birth.''107
A more recent study has appeared that compares the effect of reduced light exposure on the development of ROP in low birth weight infants. Ackerman et al. compared 161 infants whose isolettes were partially covered with a blanket to shield the eyes with 129 infants whose eyes were not shielded.109 They found no difference in the incidence or severity of ROP between the light-shielded group and the unshielded group in infants under 1,000 grams or between 1,000 and 1,500 grams. However, this study has several serious flaws. As in the Glass study, the infants were not randomized and the control group was historical. The light shielding was not begun until the infants were transferred from radiant warmers to isolettes at a median conceptual age of 30 weeks. The authors concluded that there may be "an additional benefit of transferring infants to shielded isolettes as soon as potentially possible..."109 The frequency of examinations was not standardized and some infants received only one examination. Also, many infants were lost to follow-up after discharge. While the Ackerman et al. study did not find a difference in the incidence of stage 1, 2, or 3 ROP between the shielded and unshielded groups, there were two infants (6%) who developed stage 4 ROP in the unshielded group and none in the shielded group. Again, the sample size of the study was too small to yield significant statistics. However, the study suggests that light reduction may protect against the more severe (blinding) stages of ROP. This is similar to the findings in the Glass study.107
In 1992 Seiberth et al.110 presented a poster summarizing their results of a controlled clinical trial in which 149 infants with birthweights of up to 1,500 grams were prospectively randomized to total occlusion (patching) versus no occlusion controls from Page 2-21 > 06-07-95 birth until 35 weeks of gestational age. Of the 112 infants (54 patched, 58 control) who survived until evaluation, there was no difference in the incidence of ROP between the patched group (42.6%) and the non-patched group (37.9%). Despite the prospective, randomized design of this trial, a small sample size was examined and in the up to 1000 grams birthweight group only 5 infants developed prethreshold or worse disease in both groups combined. This is hardly an adequate sample size to examine this question.
Most recently, Repka et al.111 studied a series of 160 infants with birth weights up to 1000 grams who were examined in NICUs in Baltimore and Boston. The mean illumination of the nursery in Baltimore was 463 lux and of the nursery in Boston was 126 lux. The brighter nursery (Baltimore) also had a higher incidence of any stage of ROP (64%) as compared with the dimmer nursery (41%). These results lend support to the hypothesis that light is a factor in the pathogenesis of ROP.
[Note : This comparison of the two nurseries was never published. This suppression of relevant data illustrates the medical censorship for information that could shed an unfavorable light on any hospital or doctor. One lux equals 0.093 foot-candles.]
2.7 SUMMARY AND ANALYSIS
The pathogenesis of ROP is multifactorial. The most important of these factors is the infant's maturity at the time of birth (birthweight and gestational age). However, ROP is a disease that is influenced by the interaction of environmental factors with the immature retinal vascular system of a premature infant. In the early 1950s, oxygen exposure was proven to be an extremely important factor in the exacerbation of this disease. During the past two decades, technological and therapeutic advances have been made to allow better and better control over the premature infant's oxygenation. This has made hyperoxia diminish as a variable in the production of this disease. Nonetheless, ROP still exists despite relatively good control over oxygen therapy.
[Note: Better informed nursery doctors admit among themselves that the oxygen supply of the retina is independent of the blood oxygenation level, and that no one can measure or control the oxygen levels at the retina where alone they would matter. The ignorance and misrepresentation by these authors is typical of much clinical writing on preemies.]
There are probably other environmental influences more subtle than inspired oxygen levels and duration of oxygen therapy. There is sufficient evidence to believe that precocious retinal light exposure may be such a factor. If the level of ambient light exposure is found to alter the risk of ROP, identifying this finding will provide an immediate, clinically feasible, inexpensive, non-toxic new clinical strategy to reduce the incidence and severity of this blinding disease.
> Page 2-22 > 06-07-95 Light toxicity and oxygen toxicity share a common pathogenetic mechanism and, hence, these insults may be additive. Premature infants receive high doses of ambient light and are in a biologic situation that predisposes to free radical damage. The clinical studies are contradictory and flawed. This is a solid framework that supports the theoretical rationale. Therefore, the question remains valid. In the absence of a suitable animal model in which to study light and ROP, the answer to this question can best be obtained from a well-designed clinical trial in humans. If light reduction is effective in reducing the risk of ROP, then the impact of a reduction by 20-25% could have important public health significance.
[The footnote references for this chapter are posted on page 4 in this series.]
Continue reading on page 5 more extracts from the LIGHT-ROP Manual.