1392
I. J. Radiation
Oncology
??Biology
??Physics
Finally, it is important to point out that although the method of analysis used by Wollin er al. has been shown here to be inadequate, we do not disagree with the conclusions of those authors that “. the linear quadratic model can be adopted for analysis of clinical data with results that are no worse and possibly better than the NSD model” (8). The NSD model has known limitations, especially concerning its predictions of response after short treatment times or after relatively large doses per fraction (c.f. (4), p. 222). Although the LQ model may not be valid for all tissue responses (in particular, the inclusion of a time factor is clearly necessary for some tissues), animal studies have shown that the LQ model provides a very good description of the fractionation response of certain tissues, and the model is consistent with the results of a number of clinical studies. The influence of time, dose, and fraction number on the clinical response of normal tissues certainly merits further study, but the techniques described by Wollin et al. (8) does not provide a valid method for investigating this question.
The University
1
2
Barendsen, G. W. Dose fractionation, dose rate and isoeffect relationships for normal tissue responses. Int. J. Radiaf. Oncol. Biol. Phys. 8:1981-1997; 1982. Douglas, B. G. and Fowler, J. F. The effect of multiple small doses of X rays on skin reactions in the mouse and a basic interpretation. Radiar. Res. 66:401426;
3 4. 5.
6.
7. 8
SUSANL. TUCKER Department of Biomathematics of Texas M. D. Anderson Cancer Center 1515 Holcombe Blvd. Houston, TX 77030
October
1991. Volume 21, Number 5
LQ says that two of the treatment regimens are equal (D, = 120 Gy for both), NSD says that the equivalent biological doses are 1981 and 2026 rets, and BIR calculates 1818 and 2729 (assuming five days/week fractionation). Thus, for the same 50% level of injury that was assumed for the two regimens, NSD and BIR would say that the biological effect is different. The different way the models are handling the data should be capable of analysis. That is what we attempted to do. If we assume that only one treatment regimen was used, then all the models would calculate a single biological effective dose. No model would be better than the other. It is only when the treatment regimens and effect levels change that one might see differences among the different models. The second example supposes that the LQ model will separate the injured from the non-injured, but that the isoeffective dose was in actuality incorrectly calculated. That is the luxury of examples; we assume that we know the “correct” isoeffective dose. In reality, we do not. We are only approximating what is going on with the models. We don’t know which is “correct.” We need a measure of radiation sensitivity to improve our ability to predict probability of injury from a specific dose schedule. Suppose such a measure is proposed; how shall we evaluate it? We should test its ability in retrospective studies to pick out the injured patients. We would value a new measure in accordance with its ability to provide a sharp separation of injured from noninjured. Only with such measures will we be able to deliver a tolerance dose to each patient. MYRONWOLLIN, M.S. Kaiser Permanente Medical Center Los Angeles, CA 90027
1976.
Ellis, F. Dose, time and fractionation: a clinical hypothesis. Clin. Rudiol. 20:1-7; 1969. Thames, H. D. and Hendry, J. H. Fractionation in Radiotherupy. New York: Taylor & Francis, 1987. Thames, H. D.; Peters, L. J.; Withers, H. R. and Fletcher, G. H. Accelerated fractionation vs hyperfractionation: rationales for several treatments per day. ht. J. Radiat. Oncol. Bid. Phys. 9:127-138; 1983. Thames, H. D.; Withers, H. R.; Peters, L. J. and Fletcher, G. H. Changes in early and late radiation responses with altered dose fractionation: implications for dose-survival relationships. In?. J. Radiaf. Oncol. Biol. Phys. 8:219-226; 1982. Wollin, M. and Kagan, A. R. Modification of the biologic dose to normal tissue by daily fraction. Acfa Radial. 15:481492; 1976. Wollin, M.; Kagan, A. R. and Norman, A. Predicting normal tissue injury in radiation therapy. Inf. J. Radiat. Oncol. Biol. Phys. 21(5): 1373-1376; 1991.
CRITIQUE OF “COMMENTS ON ‘PREDICTING NORMAL TISSUE INJURY IN RADIATION THERAPY’ ” To the Editor: In a Letter to the Editor, (1) Tucker criticizes the method used by Wollin et al. (2) to compare the predictions of radiobiological models. What is actually being debated is the nature of these radiobiological models. Tucker sees these models as an “important tool for estimating the total doses required to produce equal levels of normal tissue injury with changes in fractionation schedule.” While this is true, we are also assuming that these models are saying that if the value of the biological dose changes, you are at a different biological effect level and that the effect increases if the biological dose is higher. It is a common assumption in many dose-response studies. Tucker agrees that “the expected incidence of injury will be greater among patients receiving higher equivalent doses than among those receiving lower doses.” If our assumptions are correct, then one should be able to test the models for this property. In the first example, Tucker supposes a situation where the biological dose for different fractionation regimens does not change and the effect also does not change. In that situation, we should find no difference between the equivalent doses received by the injured and the non-injured patients. But in this example, all of the models should show the same results regardless of how they calculate the biological dose. They do not. While
I. Tucker,
S. L.; Comments on ‘predicting normal tissue injury in radiation therapy.’ Int. J. Radiat. Oncol. Biol. Phys. 21(5): 1391-1392; 1991. 2. Wollin, M.; Kagan, A. R.; Norman, A. Predicting normal tissue injury in radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 21(5): 1373-1376: 1991.
PAPILLARY
CYSTIC
NEOPLASM
OF THE PANCREAS
To the Ediror: Papillary cystic neoplasm (PCN) is a newly recognized tumor of the pancreas. It is distinguished from other pancreatic cancer by its early onset and favorable prognosis (4). Since Frantz described the first case in 1959, some 100 cases have been reported in the literature, most of them in young women (1). In these cases, the tumor is large, encapsulated, and hemorrhagic, presenting with abdominal pain and a palpable mass in the left upper quadrant. We present a 15.year-old previously healthy girl with 9 days history of left upper abdominal pain. CT scan and ultrasound examination showed a large cystic mass in the head of the pancreas. She underwent a whipple procedure which revealed a cystic mass at the head of the pancreas with complete obstruction of the duodenum. The tumor (9 x 5 x 5.5 cm) was soft. red-brown, encapsulated, and showed a hemorrhagic cyst measuring 6 X 6 X 2 cm. The wall of the cyst was focally disrupted in a close proximity to the posterior resection margin of the pancreas. Microscopically, the tumor showed short papillary structures composed of single or double layers of columnar cells around tine fibrovascular stalks. The tumor cells consisted of eosinophilic granular cytoplasm and peripherally located round to oval nuclei with uniformly distributed chromatin. The mitotic activity was very low. Occasional tumor cells demonstrated PAS positive and diastase resistant intracytoplasmic globules. The special stains for mucin and Grimelius stain were negative. Alcian-blue stain was weakly positive in the microcystic stroma. The sections from the area of the disrupted cyst showed tumor infiltration into the capsule, extending into the posterior resection margin of the pancreas. However, the distal pancreatic margin, small intestine, and all regional lymph nodes were free of the tumor. On immunohistochemical studies the cytoplasm of the tumor cells was diffusely positive for neuron-specific enolase and Alpha I-antitrypsin, but the stains for epithelial membrane antigen and carcinoembtyonic antigen were negative. Immunohistochemical stains for S 100, chromogranin, and other endocrine components were negative; negative for insulin, glucogen, gastrin, serotonin, somatostatin, and pancreatic polypeptide. Flow
1393
Correspondence
cytometry studies revealed that the tumor showed a DNA aneuploid pattern with a DNA index of 1.14 and a proliferative index of 10.3% (7.1% in s-phase). After surgery, the patient was given adjuvant radiotherapy at a dosage of 5040 cGy given in 28 fractions. The prognosis after surgical resection is very good. Recurrence of the tumor and metastases are very rare. Only nine cases of metastasis or recurrence have been documented (1, 3), and it is believed that the recurrence of PCN is due to incomplete resection (3). As resection was not complete in our case, it was decided to treat with adjuvant radiotherapy in an attempt to reduce the chance of local recurrence. She is alive without any signs of recurrence 6 months later. No information is available for the role of adjuvant radiotherapy in patients with incomplete resection. Pancreatic tumors have not been extensively studied using flow cytometric techniques (1, 2, 5). These studies indicate that a large percentage of pancreatic neoplasms possess an aneuploid DNA content. For pancreatic adenocarcinomas, there appears to be some correlation of a diploid DNA pattern with an improved survival and triploid DNA content to be associated with a poor survival (5). Since flow cytometric studies have been reported in two cases of PCN, little is known regarding the relevance. Both cases demonstrated readily identifiable aneuploid cell populations. The case reported by Capellari et al. (1) with a high proliferative index as determined by flow cytometry had extensive metastasis. In contrast, the case reported here showed a lower proliferative index and had no metastasis. The implications of flow cytometric results on behavior, prognosis and treatment of PCN of the pancreas remain to be elucidated.
KwH. SHIN,M.D.,F.R.C.P.(C.) Radiation Medicine Department Roswell Park Cancer Institute Buffalo, NY 14263 C. CHA M. F~KHAUSER M. PARK C. HAUER K. WEISS R. CLEVELAND
MetroHealth Medical Center Case Western Reserve University Cleveland, OH 44109
1. Cappellari, J. 0.; Geisinger, K. R.; Albertson, D. A.; Wolfman, N. T.; Kute, T. E. Malignant papillary cystic tumor of the pancreas. Cancer 66:193-198; 1990. 2. Joensuu, H.; Klemi, P. J.; Alanen, K. A. Co-existence of two aneuploid stemlines in benign adenomas. A report of three cases with stemline heterogeneity. Virchows Archiv. A Pathol. Anat. 415: 175-180; 1989. 3. Matsunou, H.; Konishi, F. Papillary-cystic neoplasm of the pancreas: a clinicopathologic study concerning the tumor aging and malignancy of nine cases. Cancer 65:283-291; 1990. 4. Toma, G. D.; Mazzocconi, G.; Campli, M.; Adami, E. A.; Gabriele, R.; Castelvetere, M.; Cesare, E. D. Papillary-cystic tumour of the pancreas. Acta Chir. Stand. 154:311-313; 1988. 5. Weger, A. R.; Falkmer, U. G.; Schwab, G. Nuclear DNA distribution pattern of the parenchymal cells in adenocarcinomas of the pancreas and in chronic pancreatitis. Gastroenterol 99:237-242; 1990.
mm will not have metastases. However, the data from Table 3 tell us that mammography cannot produce that result, since both lines in Table 3 intersect the x axis at about 20% with the highest possible frequency of examination. This implies that mammograms will not reduce the incidence of distant metastases below 20% no matter how often they are done. By interpolation of the data from Table 1, this translates to the behavior of a lesion that is 1.8 cm. This implies that the Radiological Pathology that is seen gives a necessarily limited view of the biology of the Cancer of the Breast that has been discovered and that limitation is quantifiable. Should we regard this as a fundamental bottom limit to the utility of mammography, since even with unlimited frequency of examination, the mammogram detects only a limited picture of the biological history of the cancer? Earlier diagnosis is not a statement of time as much as it is a statement of detection prior to the metastatic event. THOMAS P. O’CONNOR, M.D
Radiation Oncology Western NY Medical Park 550 Orchard Park Road West Seneca, NY 14224 1. Tubiana, M.; Koscielny, S. The natural history of breast cancer: implications for a screening strategy. Int. J. Radiat. Oncol. Biol. Phys. 19:1117-1120; 1989.
IRRADIATION OF EPITHELIAL SKIN CANCER To rhe Ediror: I read with great interest the article by Lovett et al. (1) and the accompanying editorial by Brady regarding Irradiation of epitheha1 skin cancer. The data concerning results using electron beam were of particular interest. Although electrons have been used for several decades in the treatment of these skin tumors, almost no data exist regarding their efficacy as compared to the historical data for superficial X rays. The inferior results with electrons reported by Lovett must be of concern since the availability of superficial therapeutic x-ray equipment is declining and electron irradiation is becoming common. The authors suggested several factors that might be important in obtaining optimal results with electrons, including proper use of bolus and attention to field margins. In addition, though, another important factor may be that the relative biologic effectiveness (RBE) of megavoltage electrons has been measured to be only .85 to .90 as compared to kilovoltage X rays (2). When looking at the sigmoid shaped dose response curve of epithelial tumors, this 10 to 15% reduction in biologic effectiveness may well explain the lower control rates reported by Lovett, as well as the lower incidence of moist desquamation reported by some authors with electron beam therapy (3, 4). It may be prudent, therefore, to correct the prescription of the dose and fractionation schedule to account for this when using electrons to treat tumors which have historically been treated with superficial X rays. JOHN C. BRENEMAN,M.D.
Department of Radiology Division of Radiation Oncology The Barrett Center for Cancer Prevention, Treatment and Research 234 Goodman St. Cincinnati, OH 45267-0757 1. Lovett, R. D.; Perez, C. A.; Shapiro, irradiation of epithelial skin cancer. Phys. 19:235-242; 1990.
A BOTTOM
LIMIT
TO THE UTILITY OF MAMMOGRAPHY?
To fhe Editor: While reading the lead article by Drs. Tubiana and Koscielny (1) in the November 1990 issue, I developed equations for the data in Tables 1 and 3. The data produced straight lines with the equation y = 0.67 X +0.47 describing Table 1. This was noted in the article. Table 3 has the equations y = 0.27 X - 5.3 for women between 40-49 and y = 0.25 x -5.7 for women 50-70, which are also straight lines. The implication from the first equation is that a tumor less than 4.7
S. J.; Garcia, D. M. External Int. J. Radiat. Oncol. Biol.
2. Sinclair, W. K.; Kohn, H. I. Relative Biological Effectiveness of High Energy Photons and Electrons. Radiology 82:800-814; 1964. 3. Grosch, E.; Lambert, H. E. Treatment of difficult cutaneous basal and squamous cell carcinoma with electrons. Brit. J. Radiol. 52472477; 1979. 4. Johnson, T. S.; Garciga, C. E.; Feldman, M. E., er al. Low megavoltage electron beam therapy of head and facial skin cancer using a versatile polystyrene collimeter system. Radiology 115:695-699; 1975.
I. J. Radiation
Oncology
??Biology
??Physics
Finally, it is important to point out that although the method of analysis used by Wollin er al. has been shown here to be inadequate, we do not disagree with the conclusions of those authors that “. the linear quadratic model can be adopted for analysis of clinical data with results that are no worse and possibly better than the NSD model” (8). The NSD model has known limitations, especially concerning its predictions of response after short treatment times or after relatively large doses per fraction (c.f. (4), p. 222). Although the LQ model may not be valid for all tissue responses (in particular, the inclusion of a time factor is clearly necessary for some tissues), animal studies have shown that the LQ model provides a very good description of the fractionation response of certain tissues, and the model is consistent with the results of a number of clinical studies. The influence of time, dose, and fraction number on the clinical response of normal tissues certainly merits further study, but the techniques described by Wollin et al. (8) does not provide a valid method for investigating this question.
The University
1
2
Barendsen, G. W. Dose fractionation, dose rate and isoeffect relationships for normal tissue responses. Int. J. Radiaf. Oncol. Biol. Phys. 8:1981-1997; 1982. Douglas, B. G. and Fowler, J. F. The effect of multiple small doses of X rays on skin reactions in the mouse and a basic interpretation. Radiar. Res. 66:401426;
3 4. 5.
6.
7. 8
SUSANL. TUCKER Department of Biomathematics of Texas M. D. Anderson Cancer Center 1515 Holcombe Blvd. Houston, TX 77030
October
1991. Volume 21, Number 5
LQ says that two of the treatment regimens are equal (D, = 120 Gy for both), NSD says that the equivalent biological doses are 1981 and 2026 rets, and BIR calculates 1818 and 2729 (assuming five days/week fractionation). Thus, for the same 50% level of injury that was assumed for the two regimens, NSD and BIR would say that the biological effect is different. The different way the models are handling the data should be capable of analysis. That is what we attempted to do. If we assume that only one treatment regimen was used, then all the models would calculate a single biological effective dose. No model would be better than the other. It is only when the treatment regimens and effect levels change that one might see differences among the different models. The second example supposes that the LQ model will separate the injured from the non-injured, but that the isoeffective dose was in actuality incorrectly calculated. That is the luxury of examples; we assume that we know the “correct” isoeffective dose. In reality, we do not. We are only approximating what is going on with the models. We don’t know which is “correct.” We need a measure of radiation sensitivity to improve our ability to predict probability of injury from a specific dose schedule. Suppose such a measure is proposed; how shall we evaluate it? We should test its ability in retrospective studies to pick out the injured patients. We would value a new measure in accordance with its ability to provide a sharp separation of injured from noninjured. Only with such measures will we be able to deliver a tolerance dose to each patient. MYRONWOLLIN, M.S. Kaiser Permanente Medical Center Los Angeles, CA 90027
1976.
Ellis, F. Dose, time and fractionation: a clinical hypothesis. Clin. Rudiol. 20:1-7; 1969. Thames, H. D. and Hendry, J. H. Fractionation in Radiotherupy. New York: Taylor & Francis, 1987. Thames, H. D.; Peters, L. J.; Withers, H. R. and Fletcher, G. H. Accelerated fractionation vs hyperfractionation: rationales for several treatments per day. ht. J. Radiat. Oncol. Bid. Phys. 9:127-138; 1983. Thames, H. D.; Withers, H. R.; Peters, L. J. and Fletcher, G. H. Changes in early and late radiation responses with altered dose fractionation: implications for dose-survival relationships. In?. J. Radiaf. Oncol. Biol. Phys. 8:219-226; 1982. Wollin, M. and Kagan, A. R. Modification of the biologic dose to normal tissue by daily fraction. Acfa Radial. 15:481492; 1976. Wollin, M.; Kagan, A. R. and Norman, A. Predicting normal tissue injury in radiation therapy. Inf. J. Radiat. Oncol. Biol. Phys. 21(5): 1373-1376; 1991.
CRITIQUE OF “COMMENTS ON ‘PREDICTING NORMAL TISSUE INJURY IN RADIATION THERAPY’ ” To the Editor: In a Letter to the Editor, (1) Tucker criticizes the method used by Wollin et al. (2) to compare the predictions of radiobiological models. What is actually being debated is the nature of these radiobiological models. Tucker sees these models as an “important tool for estimating the total doses required to produce equal levels of normal tissue injury with changes in fractionation schedule.” While this is true, we are also assuming that these models are saying that if the value of the biological dose changes, you are at a different biological effect level and that the effect increases if the biological dose is higher. It is a common assumption in many dose-response studies. Tucker agrees that “the expected incidence of injury will be greater among patients receiving higher equivalent doses than among those receiving lower doses.” If our assumptions are correct, then one should be able to test the models for this property. In the first example, Tucker supposes a situation where the biological dose for different fractionation regimens does not change and the effect also does not change. In that situation, we should find no difference between the equivalent doses received by the injured and the non-injured patients. But in this example, all of the models should show the same results regardless of how they calculate the biological dose. They do not. While
I. Tucker,
S. L.; Comments on ‘predicting normal tissue injury in radiation therapy.’ Int. J. Radiat. Oncol. Biol. Phys. 21(5): 1391-1392; 1991. 2. Wollin, M.; Kagan, A. R.; Norman, A. Predicting normal tissue injury in radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 21(5): 1373-1376: 1991.
PAPILLARY
CYSTIC
NEOPLASM
OF THE PANCREAS
To the Ediror: Papillary cystic neoplasm (PCN) is a newly recognized tumor of the pancreas. It is distinguished from other pancreatic cancer by its early onset and favorable prognosis (4). Since Frantz described the first case in 1959, some 100 cases have been reported in the literature, most of them in young women (1). In these cases, the tumor is large, encapsulated, and hemorrhagic, presenting with abdominal pain and a palpable mass in the left upper quadrant. We present a 15.year-old previously healthy girl with 9 days history of left upper abdominal pain. CT scan and ultrasound examination showed a large cystic mass in the head of the pancreas. She underwent a whipple procedure which revealed a cystic mass at the head of the pancreas with complete obstruction of the duodenum. The tumor (9 x 5 x 5.5 cm) was soft. red-brown, encapsulated, and showed a hemorrhagic cyst measuring 6 X 6 X 2 cm. The wall of the cyst was focally disrupted in a close proximity to the posterior resection margin of the pancreas. Microscopically, the tumor showed short papillary structures composed of single or double layers of columnar cells around tine fibrovascular stalks. The tumor cells consisted of eosinophilic granular cytoplasm and peripherally located round to oval nuclei with uniformly distributed chromatin. The mitotic activity was very low. Occasional tumor cells demonstrated PAS positive and diastase resistant intracytoplasmic globules. The special stains for mucin and Grimelius stain were negative. Alcian-blue stain was weakly positive in the microcystic stroma. The sections from the area of the disrupted cyst showed tumor infiltration into the capsule, extending into the posterior resection margin of the pancreas. However, the distal pancreatic margin, small intestine, and all regional lymph nodes were free of the tumor. On immunohistochemical studies the cytoplasm of the tumor cells was diffusely positive for neuron-specific enolase and Alpha I-antitrypsin, but the stains for epithelial membrane antigen and carcinoembtyonic antigen were negative. Immunohistochemical stains for S 100, chromogranin, and other endocrine components were negative; negative for insulin, glucogen, gastrin, serotonin, somatostatin, and pancreatic polypeptide. Flow
1393
Correspondence
cytometry studies revealed that the tumor showed a DNA aneuploid pattern with a DNA index of 1.14 and a proliferative index of 10.3% (7.1% in s-phase). After surgery, the patient was given adjuvant radiotherapy at a dosage of 5040 cGy given in 28 fractions. The prognosis after surgical resection is very good. Recurrence of the tumor and metastases are very rare. Only nine cases of metastasis or recurrence have been documented (1, 3), and it is believed that the recurrence of PCN is due to incomplete resection (3). As resection was not complete in our case, it was decided to treat with adjuvant radiotherapy in an attempt to reduce the chance of local recurrence. She is alive without any signs of recurrence 6 months later. No information is available for the role of adjuvant radiotherapy in patients with incomplete resection. Pancreatic tumors have not been extensively studied using flow cytometric techniques (1, 2, 5). These studies indicate that a large percentage of pancreatic neoplasms possess an aneuploid DNA content. For pancreatic adenocarcinomas, there appears to be some correlation of a diploid DNA pattern with an improved survival and triploid DNA content to be associated with a poor survival (5). Since flow cytometric studies have been reported in two cases of PCN, little is known regarding the relevance. Both cases demonstrated readily identifiable aneuploid cell populations. The case reported by Capellari et al. (1) with a high proliferative index as determined by flow cytometry had extensive metastasis. In contrast, the case reported here showed a lower proliferative index and had no metastasis. The implications of flow cytometric results on behavior, prognosis and treatment of PCN of the pancreas remain to be elucidated.
KwH. SHIN,M.D.,F.R.C.P.(C.) Radiation Medicine Department Roswell Park Cancer Institute Buffalo, NY 14263 C. CHA M. F~KHAUSER M. PARK C. HAUER K. WEISS R. CLEVELAND
MetroHealth Medical Center Case Western Reserve University Cleveland, OH 44109
1. Cappellari, J. 0.; Geisinger, K. R.; Albertson, D. A.; Wolfman, N. T.; Kute, T. E. Malignant papillary cystic tumor of the pancreas. Cancer 66:193-198; 1990. 2. Joensuu, H.; Klemi, P. J.; Alanen, K. A. Co-existence of two aneuploid stemlines in benign adenomas. A report of three cases with stemline heterogeneity. Virchows Archiv. A Pathol. Anat. 415: 175-180; 1989. 3. Matsunou, H.; Konishi, F. Papillary-cystic neoplasm of the pancreas: a clinicopathologic study concerning the tumor aging and malignancy of nine cases. Cancer 65:283-291; 1990. 4. Toma, G. D.; Mazzocconi, G.; Campli, M.; Adami, E. A.; Gabriele, R.; Castelvetere, M.; Cesare, E. D. Papillary-cystic tumour of the pancreas. Acta Chir. Stand. 154:311-313; 1988. 5. Weger, A. R.; Falkmer, U. G.; Schwab, G. Nuclear DNA distribution pattern of the parenchymal cells in adenocarcinomas of the pancreas and in chronic pancreatitis. Gastroenterol 99:237-242; 1990.
mm will not have metastases. However, the data from Table 3 tell us that mammography cannot produce that result, since both lines in Table 3 intersect the x axis at about 20% with the highest possible frequency of examination. This implies that mammograms will not reduce the incidence of distant metastases below 20% no matter how often they are done. By interpolation of the data from Table 1, this translates to the behavior of a lesion that is 1.8 cm. This implies that the Radiological Pathology that is seen gives a necessarily limited view of the biology of the Cancer of the Breast that has been discovered and that limitation is quantifiable. Should we regard this as a fundamental bottom limit to the utility of mammography, since even with unlimited frequency of examination, the mammogram detects only a limited picture of the biological history of the cancer? Earlier diagnosis is not a statement of time as much as it is a statement of detection prior to the metastatic event. THOMAS P. O’CONNOR, M.D
Radiation Oncology Western NY Medical Park 550 Orchard Park Road West Seneca, NY 14224 1. Tubiana, M.; Koscielny, S. The natural history of breast cancer: implications for a screening strategy. Int. J. Radiat. Oncol. Biol. Phys. 19:1117-1120; 1989.
IRRADIATION OF EPITHELIAL SKIN CANCER To rhe Ediror: I read with great interest the article by Lovett et al. (1) and the accompanying editorial by Brady regarding Irradiation of epitheha1 skin cancer. The data concerning results using electron beam were of particular interest. Although electrons have been used for several decades in the treatment of these skin tumors, almost no data exist regarding their efficacy as compared to the historical data for superficial X rays. The inferior results with electrons reported by Lovett must be of concern since the availability of superficial therapeutic x-ray equipment is declining and electron irradiation is becoming common. The authors suggested several factors that might be important in obtaining optimal results with electrons, including proper use of bolus and attention to field margins. In addition, though, another important factor may be that the relative biologic effectiveness (RBE) of megavoltage electrons has been measured to be only .85 to .90 as compared to kilovoltage X rays (2). When looking at the sigmoid shaped dose response curve of epithelial tumors, this 10 to 15% reduction in biologic effectiveness may well explain the lower control rates reported by Lovett, as well as the lower incidence of moist desquamation reported by some authors with electron beam therapy (3, 4). It may be prudent, therefore, to correct the prescription of the dose and fractionation schedule to account for this when using electrons to treat tumors which have historically been treated with superficial X rays. JOHN C. BRENEMAN,M.D.
Department of Radiology Division of Radiation Oncology The Barrett Center for Cancer Prevention, Treatment and Research 234 Goodman St. Cincinnati, OH 45267-0757 1. Lovett, R. D.; Perez, C. A.; Shapiro, irradiation of epithelial skin cancer. Phys. 19:235-242; 1990.
A BOTTOM
LIMIT
TO THE UTILITY OF MAMMOGRAPHY?
To fhe Editor: While reading the lead article by Drs. Tubiana and Koscielny (1) in the November 1990 issue, I developed equations for the data in Tables 1 and 3. The data produced straight lines with the equation y = 0.67 X +0.47 describing Table 1. This was noted in the article. Table 3 has the equations y = 0.27 X - 5.3 for women between 40-49 and y = 0.25 x -5.7 for women 50-70, which are also straight lines. The implication from the first equation is that a tumor less than 4.7
S. J.; Garcia, D. M. External Int. J. Radiat. Oncol. Biol.
2. Sinclair, W. K.; Kohn, H. I. Relative Biological Effectiveness of High Energy Photons and Electrons. Radiology 82:800-814; 1964. 3. Grosch, E.; Lambert, H. E. Treatment of difficult cutaneous basal and squamous cell carcinoma with electrons. Brit. J. Radiol. 52472477; 1979. 4. Johnson, T. S.; Garciga, C. E.; Feldman, M. E., er al. Low megavoltage electron beam therapy of head and facial skin cancer using a versatile polystyrene collimeter system. Radiology 115:695-699; 1975.
