Cancers related to contraceptive use.

Sir In their letter Poljak et al. expressed their experience in processing of archival cervical smears for human papillomavirus (HPV) detection by polymerase chain reaction (PCR). In agreement with our observation, they found that a complete DNA extraction procedure is required for a successful amplification by PCR. However, they claim that DNA of sufficient molecular weight can easily be isolated to allow amplification of DNA fragments longer than 200 bp. This is in contrast to our findings, showing that it is far from easy to extract sufficient amounts of larger DNA fragments, particularly from older smears. It should be noted that this discrepancy may reflect differences in preparation of the Papanicolaou smear. In particular, the fixation procedure used might be of importance, as the preservation of nucleic acids is likely to be fixation dependent. It is well known that among the different cytological laboratories different fixatives are used to ensure an optimal morphology. This was discussed in our paper. Obviously, the fact that Poljak et al. were able to extract relatively large DNA fragments from their older smears reflects a better preservation of nucleic acids than in the long-stored smears analysed in our laboratory. However, our studies also revealed long-stored smears from which larger fragments could be amplified, indicating that the correlation between storage time and mean fragment size is not strictly linear. Because of the lack of proper documentation, we do not know if some of the long-stored smears analysed in our study were originally fixed in another way, which could be an alternative explanation for the observed differences. Concerning the complexity of the GTC/silica beads method, we have the impression that, based on recent experiments, the use of diatoms rather than silica particles improves the yield after one extraction round. The use of diatoms has been described as one of the alternatives in the GTC/silica protocol (Boom et al., 1990). Currently, we have no explanation for this observation, but since diatoms are much larger than size-fractionated silica particles, it is conceivable that diatoms display less stringent requirements for nucleic acids binding. Although we did not test the -30°C/37°C alternate incubation protocol for removal of coverslips, this suggestion is of interest and may save time. In conclusion, before definitive conclusions can be drawn about the feasibility of different extraction methods, the effect of fixatives used for preparing Pap smears has to be analysed.


Reply to the letter from Poljak et al
Sir -In their letter Poljak et al. expressed their experience in processing of archival cervical smears for human papillomavirus (HPV) detection by polymerase chain reaction (PCR). In agreement with our observation, they found that a complete DNA extraction procedure is required for a successful amplification by PCR. However, they claim that DNA of sufficient molecular weight can easily be isolated to allow amplification of DNA fragments longer than 200 bp. This is in contrast to our findings, showing that it is far from easy to extract sufficient amounts of larger DNA fragments, particularly from older smears. It should be noted that this discrepancy may reflect differences in preparation of the Papanicolaou smear. In particular, the fixation procedure used might be of importance, as the preservation of nucleic acids is likely to be fixation dependent. It is well known that among the different cytological laboratories different fixatives are used to ensure an optimal morphology. This was discussed in our paper. Obviously, the fact that Poljak et al. were able to extract relatively large DNA fragments from their older smears reflects a better preservation of nucleic acids than in the long-stored smears analysed in our laboratory. However, our studies also revealed long-stored smears from which larger fragments could be amplified, indicating that the correlation between storage time and mean fragment size is not strictly linear. Because of the lack of proper documentation, we do not know if some of the long-stored smears analysed in our study were originally fixed in another way, which could be an alternative explanation for the observed differences.
Concerning the complexity of the GTC/silica beads method, we have the impression that, based on recent experiments, the use of diatoms rather than silica particles improves the yield after one extraction round. The use of diatoms has been described as one of the alternatives in the GTC/silica protocol (Boom et al., 1990). Currently, we have no explanation for this observation, but since diatoms are much larger than size-fractionated silica particles, it is conceivable that diatoms display less stringent requirements for nucleic acids binding.
Although we did not test the -30°C/37°C alternate incubation protocol for removal of coverslips, this suggestion is of interest and may save time.
In conclusion, before definitive conclusions can be drawn about the feasibility of different extraction methods, the effect of fixatives used for preparing Pap smears has to be analysed.

Cancers related to contraceptive use
Sir -There is another hypothesis in addition to contraceptive use or oestrogen replacement therapy to account for the changing gender ratios of colorectal cancer reported by Silva and Swerdlow (1996). In their discussion of declining F/M incidence ratios, the authors overlooked the contribution that non-steroidal anti-inflammatory drugs (NSAIDs) may have.
NSAIDs inhibit prostaglandin synthesis, which is elevated in human colorectal cancers, in venous blood therefrom and in peripheral blood in patients with metastases (Marnett, 1992;Narisawa et al., 1990). NSAID therapy reduced the incidence of carcinogen-induced large bowel neoplasms in rodents, unlike oral contraceptives or oestrogens. Aspirin and newer NSAIDs have delayed recurrence of polyps in hereditary polyposis of the colon and reduced by 30-50% colorectal cancer development in case-control and prospective clinical trials (Giardiello et al., 1995;Giovannucci et al., 1995).
Salicylates have been available since early in this century, and newer synthetic NSAIDs have been widely used for pain relief only during the past 30 years. The latter have been very profitable and hence heavily advertised. Not counting overthe-counter sales, well over 70 000 000 prescriptions for these drugs are filled annually in the US. In addition to premenstrual symptoms, women have about 60% more headaches, arthritis and back pains than men (Benson and Marano, 1994). Women also have more sluggish colonic peristalsis than men, perhaps increasing their cancer risk (Lampe et al., 1993).
Since 1940, colorectal cancer mortality rates (age adjusted to the 1970 US population) in US women has declined with little change in men; the F/M ratio fell from 1.0 to 0.68 (Wingo et al., 1995;Page and Asire, 1985;Riss et al., 1994). Since 1970, the Nesbraska ratio has declined by over 40%, falling from 0.93 to 0.53, resulting in 200+ fewer deaths annually. From 1987 to 1993, there were no significant gender differences in staging or operability of these tumours in our state (Nebraska Dept. of Health, 1991).
Consumer surveys are needed to determine which of these therapies may be most important in contributing to this welcome decrease in the incidence and mortality of colorectal cancer in women, but not yet in men.

Reply to the letter from Rettig and Lemon
Sir -In their letter, Rettig and Lemon raised the possibility that non-steroidal anti-inflammatory drugs (NSAIDs) may have contributed to the sex differences in time trends of colorectal cancer in England and Wales. We agree that this is an interesting possibility; there are some data available to examine it. The total number of prescriptions (excluding preparations sold over the counter) for NSAIDs other than aspirin increased markedly in the United Kingdom from 7.6 million in 1967 to 22 million in 1985(Walt et al., 1986. The increase occurred in both sexes but prescription rates were 40-60% higher in women than in men throughout the period. Data on aspirin consumption are available from a nationally representative survey carried out in Great Britain in 1973(OPCS, 1976. In this survey, women were 36% more likely than men to have consumed any analgesics containing aspirin in the week preceding the survey (Table I); this female excess was particularly marked at ages 45-64 years. Among consumers, women were also more likely to have taken these drugs more often: 58% of women had taken these drugs at least twice a week, whereas the corresponding figure for men was 48%.
The increase in NSAID consumption may have contributed to the decrease in the risk of colorectal cancer in women in England and Wales in recent years. However, the consumption of these drugs has also increased markedly in men (although less than in women) but no decline in their cancer risks has yet been observed. The few analytical studies conducted so far with sex-specific data do not seem to indicate that the effect of these drugs on the risk of developing colorectal cancer may be greater in women than in men (Muscat et al., 1994;Rosenberger et al., 1991).