Part 1. The Early Reliance on X-Ray Therapy of Acute Inflammations
In the years before the introduction of chemotherapeutic agents such as the sulfonamides and the antibiotics such as penicillin, there existed a large, developed reliance on x-ray therapy of acute inflammations. Furuncles, carbuncles, cellulitis, pneumonias, erysipelas, and many other acute inflammatory diseases were treated with x-rays. This was true for some inflammatory states other than those caused by acute infections as well as those with an infectious origin.
It may seem to some, today, that these were poor uses of x-ray therapy. Hindsight is so available. However, putting oneself back into those years before chemotherapy and antibiotics --- with the terrifying inability to cope with acute inflammatory conditions --- may lead to a more sympathetic view of such uses.
We shall later discuss (Chapter 18) some studies of acute pneumonias successfully treated with x-rays, after sulfonamides (ordinarily successful) had failed badly.
Why X-Ray Therapy
for Acute Postpartum Mastitis Was Delayed in Acceptance
In 1945, Roger Harvey and colleagues of the Strong Memorial Hospital in Rochester, New York, wrote a paper providing some of the useful historical details on the introduction of x-ray therapy in the treatment of acute postpartum mastitis --- a troubling condition vividly described in Part 2.
The reader might wonder why there were not comprehensive reports on this particular application of x-ray therapy at some earlier date. Harvey and colleagues believed they had an answer to this question.
They related that roentgen treatment of acute postpartum mastitis had been successful in Europe, and they listed reports by R. Goetz (1939), J. Granzo (1931), R. Goedel (1937), W.H. Hanne (1936), K. Kautsky (1934), C. Margraf (1936), G.J. Pfalz (1934), A. Pohl (1939), E. Steinkamm (1936), H. Theiss (1935). Also, they related that it had been successful in the few clinics from which it had been reported in the United States, and they listed papers by J.F. Elward and S.M. Dodek (1940), and by H.C. McIntosh (1940).
Harvey and co-workers suggested that widespread acceptance of this method of treatment had been hampered by several specific failings, which they listed (p.396):
(1) Failure to realize that a co-ordinated plan of management of these patients is necessary even when roentgen therapy is given;
(2) Failure to prevent simultaneous combinations of conflicting forms of treatment;
(3) Delay in referring these patients for roentgen therapy until other methods of treatment had failed to relieve the patient.
And they related that in January, 1942, a co-operative plan of treatment of acute mastitis was undertaken by the departments of obstetrics and radiology at their hospital. Since that time, the only breast abscesses which occurred were in some of the patients initially treated at home and subsequently referred for roentgen treatment, or in patients treated entirely by methods other than roentgen therapy.
For those unfamiliar with such matters, abscess formation represents a failure of treatment, in that surgical drainage is then required to manage this complication.Part 2. Harvey's Important Words
on the Successful Therapy and Its Mystery
In the 1945 paper, Dr. Harvey and colleagues described their methods and experience with roentgen therapy in treating 100 patients with acute mastitis between January of 1942 and June of 1944. This number of patients represented a very aggressive increase in the use of roentgen therapy for this disease in the Strong Memorial Hospital, inasmuch as the paper stated that, in the preceding 16 years of the hospital's existence, a total of only 77 patients had been treated for acute postpartum mastitis by all methods combined.
A Brief Description of the Diagnosis of Acute Mastitis
We quote from Harvey 1945 (p.396):
"The diagnosis of this disease is most accurately made by a combination of subjective symptoms and objective observations. Some patients note pain from cracked or fissured nipples from one to several nursing periods before the onset of mastitis. Then a portion of the breast becomes tender, and fever, aches, and chills occur. Other patients have an explosive onset of their symptoms with high fever, chills, aches, sweats, headache, coryza, and what the patients consider or admit to be a little incidental tenderness of one breast. In the latter group, urine cultures, uterine cultures, chest roentgenograms, and ear, nose, and throat examinations are sometimes in progress before it is realized that a breast infection could account for all the symptoms."
A Comparison of Women Treated and Untreated with X-Ray Therapy
Dr. Harvey's paper compares women treated and untreated with x-rays (p.400):
"The literature on roentgen treatment of acute postpartum mastitis leads one to believe that striking relief from all symptoms occurs immediately after treatment. We have never observed this in treating other infections such as furuncles, parotitis and cellulitis and have not been surprised to find that the same type of interval exists between treatment and effect in this condition. It is granted that a few of our patients have been free from all symptoms two hours after treatment, but in any large series of patients there will be some mild infections which will subside quickly irrespective of treatment. Most of the roentgen treated patients actually do go through some type of crisis 4 to 8 hours after treatment, and thereafter do feel markedly improved. Careful tabulation of the maximum duration of all subjective symptoms and objective findings in the roentgen treated series reveals that the average duration of subjective symptoms is 1.88 days, and objective findings, 2.45 days. In contrast are the figures of the group not treated with roentgen rays in whom suppuration did not occur, with an average of 8 days of subjective symptoms and 10.3 days of objective findings. The group operated on had subjective symptoms from infection for an average of 44.6 days and objective findings for an average of 47.7 days per patient." And:
"No one knows definitely what reaction takes place in an irradiated infection of this type. We do know that the doses of roentgen rays used are infinitesimal as far as ability directly to affect the bacteria themselves."
We believe that this oft-reported finding, of clinical response in infections with doses that could not conceivably prove bacteriostatic or bacteriocidal, remains a mystery today. Harvey and co-workers (p.401) cite Desjardins' suggestion of a speedup of the whole process of defense against local infection.
Harvey's Bottom Line (p.402): Treat Promptly, "Regardless of the Hour"
"Application of roentgen therapy is so effective in the early stages of this disease that treatment should be started as soon as the diagnosis can be definitely made, regardless of the hour. Delay in starting this treatment decreases the chances of recovery without surgery."Part 3. Analysis of the Quantitative Details
in a Much Larger Rochester Series
In developing our dose-entries for the Master Table, Column E, we will use data from a much larger series than Dr. Harvey's. We will begin with the 1977 paper by Dr. Roy E. Shore, Dr. Louis Hempelmann, and others, entitled "Breast Neoplasms in Women Treated with X-Rays for Acute Postpartum Mastitis." We will follow a "checklist" of items similar to the model in Chapter 8.
- Item 1: What was the place of study? Rochester, New York, in Monroe County. We are told the following in the Shore Study:
"The 606 women in group A were treated for acute postpartum mastitis with X-rays by Rochester radiologists largely between 1940 and 1955, but a few were treated later. Although X-rays were used to treat acute postpartum mastitis with varying degrees of success, local obstetricians and radiologists claimed the treatment was highly successful if given early. The treatment was considered an emergency procedure and was given day or night as soon as the diagnosis was made."
- Item 2: Can we regard the listed participants as truly representative of Monroe County for the relevant period? The answer is, "Not sure."
The treatments were done by Rochester radiologists, but we do not have it spelled out that all cases treated in Monroe County were included. That is, the Shore Study does not tell us that the coverage was essentially complete, as it was in Chapter 8 for the infants treated for enlarged thymus disease. So, we must leave open the possibility that the true number of persons treated, and hence the true person-rads delivered to breasts, may be underestimated for this series in Monroe County. This means we would also underestimate the nationwide person-rads in the Master Table, Column E.
- Item 3: How many persons were treated? In the Shore Study, there were 606 women who received the radiation therapy.
- Item 4: What ages were represented in the treated group? We can approximate that there were 31 different age-years represented. All except one woman were in the age range of 14-44 years at treatment, and the majority were 20-34 years of age (see Item 6).
- Item 5: What was the total period over which the treatments were performed? According to Dr. Shore and co-workers, the treatments were "largely" between 1940 and 1955, but a few were treated later. Other clinics were using the method well before the Rochester study. We shall use an adjusted value of 20 years out of 40 for duration of such use.
Females of Each Age-Year Irradiated per Calendar-Year, Monroe County
- Item 6: We need to know how many women of each age-year were treated during one average year of the study. This is a 3-step process which produces the next tabulation.
The first step is to divide the 606 cases into age-bands according to the percentages provided by the Shore Study. Column B of the tabulation shows the percentages, and Column C shows the numbers of cases per age-band.
The second step is to divide each number in Column C by 15, because the cases were accumulated over a period of 15 years (Item 5). The result in Column D is the average number of cases treated in one calendar-year. The Master Table always considers a single year.
The third step is to obtain the average number of cases treated in each age-year. So we divide the cases per age-band (in Column D) by the number of age-years in the band (Column E). And thus we arrive at a reasonable estimate of cases treated in each age-year during one calendar-year, in Column F of the tabulation.
Age- Percent Cases Cases/15 Size of Cases per Band of Total Band Age-Year A B C D E F 14-19 4 % 24.24 1.62 6 0.27 20-24 27 % 163.62 10.91 5 2.18 25-29 39 % 236.34 15.76 5 3.15 30-34 20 % 121.20 8.08 5 1.62 35-39 8 % 48.48 3.23 5 0.65 40-44 2 % 12.12 0.81 5 0.16
Females of Age-Years 14-44, per Year, USA and Monroe County
- Item 7: We need to know the total population of females in each age-year (14-44) in Monroe County for the average year in the period 1920-1960. We will make the estimate with the same method used in Chapter 8 (Item 6) and Chapter 9 (Item 7). We begin with the ratio of the Monroe County population to the U.S. population in 1960, from Chapter 8, Item 6:
The ratio, Monroe / USA, was (586,309 / 179,333,000), or 0.00327.
Then we find the national number of women per age-year (age-years 14-44) in the Master Table, Column A, and we multiply each national number by the Monroe / USA ratio to obtain the appropriate estimates for Monroe County, which are shown in the next tabulation. Since the national numbers of women do not change very much in a five-year interval, we are using one value for an entire age-band.
Age-Band National Monroe County Number per Age-Year Number per Age-Year 14-19 887,609 2,902 20-24 884,794 2,893 25-29 881,805 2,884 30-34 878,278 2,872 35-39 873,473 2,856 40-44 860,227 2,813
Breast-Dose per Treated Woman
- Item 8: We need the average dose to breast-pairs from the therapy. Shore and co-workers give the following mean breast-pair doses (in medical rads), and they are so similar that we can take an unweighted average.
Treated at age-years 15-29, mean breast-pair dose = 251 medical rads.
Treated at age-years 30-35, mean breast-pair dose = 239 medical rads.
Mean dose, without any adjustment for supra-linearity = 245 rads / breast-pair. Adjustment for supra-linearity = factor of 0.398 (Chapter 8, Box in Item 8). Adjusted mean dose = 97.5 medical rads to the breast-pair.
Conversion of Individual Dose to Population-Dose
- Item 9: We need the average population-dose from this therapy, rather than the raw individual dose per treatment. This is a two-step process, in which we obtain the person-rads per age-year, and then divide person-rads by the total persons in the county. Because there are so many age-years involved, we will need to make another tabulation.
For the new tabulation, Item 6 provides the information for Column B: Number of women treated per age-year. And Item 8 provides the breast-dose per treated woman for Column C. So Column D, which is person-rads, is Col.B times Col.C.
The total persons (female) in the county comes into Column E from Item 7. By dividing person-rads (Col.D) by persons (Col.E), we obtain the average Population-Dose per Age-Year for Column F. The final adjustment seen in Column G comes from Item 10, below.
Treated Total Pop'n Adjust Age- Women Dose in Person- Women Dose per for Band, per Medical Rads in Each Age-Year 20 Yrs Years Age-Year Rads Age-Year A B C D E F G 14-19 0.27 97.5 26.33 2,902 0.0091 0.0045 20-24 2.18 97.5 212.71 2,893 0.0735 0.0368 25-29 3.15 97.5 307.24 2,884 0.1065 0.0533 30-34 1.62 97.5 157.56 2,872 0.0549 0.0274 35-39 0.65 97.5 63.02 2,856 0.0221 0.0110 40-44 0.16 97.5 15.60 2,813 0.0055 0.0028
After one more adjustment in Item 10, this tabulation will provide 31 separate entries into the Master Table, since there are 31 age-years represented here (females, 14 through 44 years).
- Item 10: Duration. The last adjustment, already seen in Column G of the preceding tabulation, is related to the 20 years out of all 40 years (1920-1960) that we should "credit" with this source of radiation. This represents a slight increment in duration over the Rochester study itself, as indicated in Item 5, to take into account the earlier experience of other clinics. In view of extensive use in Europe in the 1930s, it is not conceivable that no American clinics were using the method. Indeed, we have American reports in the year 1940, which had to reflect work done before 1940, as does Desjardins 1931. The entries in Column G above are transferred to our Master Table, Column E. These may be underestimates because of uncertainty in Item 2.
# # # # #
Excerpts (Part 1) from an Important Paper of 1931:
"Radiotherapy for Inflammatory Conditions"
by an eminent roentgenological scholar of the era,
Arthur U. Desjardins, M.D., of the Mayo Clinic, Rochester, Minnesota.
In the Journal of the American Medical Association: 96, No.6: 401-408. Feb. 7, 1931.
The first paragraph follows: