- Email: [email protected]
Blood Dyscrasias and Carbonic Anhydrase Inhibitors
Blood Dyscrasias and Carbonic Anhydrase Inhibitors LYLAS GOOD MOGK, MD,l MARSHALL N. CYRLIN, MD 2
Abstract: Two new cases of aplastic anemia possibly associated with Neptazane (methazolamide) are reported. Several previous authors, as well as the manufacturer of Diamox (acetazolamide) and Neptazane, have recommended routine blood counts for patients on carbonic anhydrase inhibitors. Four surveys are presented of current practices with regard to blood monitoring. These include authors of case reports, 40 academic ophthalmologists, 81 ophthalmologists in private practice, and 66 glaucoma specialists. The vast majority does not routinely monitor blood counts of patients on carbonic anhydrase inhibitors. The value of routine blood monitoring is questionable both because of (1) the idiosyncratic, non-dose-related mechanism of the dyscrasias and (2) the variability of the timing of their onset and development. Regular observation and questioning of patients for symptoms are thought to be preferable. The importance of a thorough history when assigning an etiology to a dyscrasia is noted. [Key words: acetazolamide, aplastic anemia, blood dyscrasias, carbonic anhydrase inhibitor, Diamox, methazolamide, Neptazane.] Ophthalmology 95:768771' 1988
As the number of case reports of blood dyscrasias possibly associated with carbonic anhydrase inhibitors increases, the response of the authors has often been to encourage routine blood counts for early detection. Two cases presented herein beg the question as to whether such blood counts could have detected the dyscrasias and helped prevent the fatal outcome. To answer this question, we reviewed the current knowledge of the mechanism and pattern of carbonic anhydrase inhibitor-induced dyscrasias; we surveyed monitoring practices of ophthalmologists, previous authors, ophthalmologic academicians, and glaucoma specialists. In this article we present their practices and recommendations for monitoring patients on carbonic anhydrase inhibitors.
CASE REPORTS Case 1. A 71-year-old white man was first seen on May 24, 1984. He was using Carbachol twice daily and had visual Originally received: July 13, 1987. Revision accepted: January 7, 1988. 1
2
Kresge Eye Institute and Grosse Pointe Ophthalmology PC, Detroit. Franklin Eye Consultants, Detroit.
Reprint requests to Lylas Good Mogk, MD, Grosse Pointe Ophthalmology PC, 15401 E. Jefferson Ave, Grosse Pointe Park, Ml 48230.
768
acuity of 20/60 in the right eye and 20/30 in the left, intraocular pressures (lOPs) of 33 and 24 mmHg, and cup: disc ratios of 0.9 and 0.7. Carbachol was increased to three times daily and Timoptic 0.5% and Propine® 0.1% were added. On July 7, his visual acuity was 20/100 in the right eye (lOP, 23 mmHg) and 20/40 in the left (lOP, 21 mmHg), and he was started on Neptazane 50 mg twice daily. He was seen monthly. On September 21, with lOPs of 24 mmHg in the right eye and 19 mmHg in the left, an argon laser trabeculoplasty was performed on the right eye. On October 6, his lOPs were 23 mmHg in the right eye and 18 mmHg in the left. On October 12, the patient phoned to report weakness in his legs. He was instructed to discontinue Neptazane and come in. He did discontinue Neptazane on October 12 but did not return to the ophthalmologist. On October l 7, he was admitted by his internist to a local hospital with fatigue, chills, sore throat, lower extremity ecchymosis, and a history of a 13-lb weight loss over the preceding 3 months. The patient was not on medication, had no history of sulfonamide allergy, and no history of chemical or environmental exposures was elicited. Laboratory results were as follows: hemoglobin 5.5 g/dl; leukocyte count 1500 Ill with 99% lymphocytes; and platelets, 2000. Bone marrow biopsy done shortly after his October 17 admission showed acellular damaged marrow with fatty replacement, and the diagnosis of aplastic anemia was made. Throat cultures grew Staphylococcus aureus. Over the ensuing 2 weeks, the patient was treated with multiple antibiotics including Tobramycin, Ticarcillin, Vancomycin, amphotericin B, Cefabid, and Amikacin. Multiple transfusions of erythrocytes and platelets were given, the hemoglobin level rose to 8.6 g/dl, leukocyte count to 14,000 Ill, and platelet count to
MOGK AND CYRLIN
•
BLOOD DYSCRASIAS AND CAis
Table 1. Academicians Polled Regarding Blood Count Monitoring of Patients on Carbonic Anhydrase Inhibitors Institution Bascom Palmer Beth Israel, Boston Kresge Eye Institute Louisiana State University Massachusetts Eye and Ear Mt. Sinai, New York New York-Cornell New York University Ochsner Clinic Scheie Institute Sinai Hospital, Detroit Tufts New England Tulane University University of California at San Francisco University of Colorado Wills Eye Institute Wilmer Institute Total *
No. Contacted
No. Responding
No. Who Monitor
4* 1 2* 1 2* 2* 2* 1 3* 3 2* 3 3 2* 1 3 3*
0 0 0 1 1t 0 0 0 0 0 0 0 1 0 0 1:j: 0
5
1 2
5
4 2 2 3
5
4 2 4
5
2 1
5
4
40
56
Time Interval
Before treatment
Initially and every 3 mos Every 3 mos
4
Includes glaucoma specialist(s).
t Only if symptoms, not routinely.
:j: To monitor disease processes unrelated to carbonic anhydrase inhibitors.
59,000. On November 7, the patient's temperature was 105"F, and on November 9 he died of septic shock. No autopsy was performed. The death certificate cited methazolamide as the cause of the aplastic anemia. Case 2. A 73-year-old white woman was first seen on May 10, 1984, with a visual acuity of 20/30 in the right eye and 20/15 in the left, lOPs of24 mmHg in both eyes, and enlarged optic cups. She was on Timoptic 0.5% twice daily and on no other medications. Propine 0.1% was added, and 1 month later her lOPs were 23 mmHg in both eyes. Propine was replaced by Pilocarpine 1%, then 2%, and finally 4%. On September 17, with a visual acuity of 20/40 in the right eye and 20/25 in the left and lOPs of23 mmHg in both eyes, the patient was started on Neptazane 50 mg twice daily, and Pilocarpine was discontinued. On October 10, lOPs were 21 mmHg in the right eye and 20 mmHg in the left, and Neptazane was increased to three times daily. Argon laser trabeculoplasties were performed on October 15 and 29, and on October 31 lOPs were 16 mmHg in both eyes, and the patient had no complaints. She was continued on Neptazane. On November 13, the patient complained to her internist of diarrhea, for which she was taking Kaopectate. She was given lmodium, and a complete blood count was ordered which showed a leukocyte count of 3600 ~1 with 39 polymorphonucleocytes, 55 lymphocytes, 6 monocytes, and no bands; her platelet count was 257,000. The decreased leukocyte count did not arouse suspicion. The patient was seen by her ophthalmologist on December 5, with lOPs of 22 mmHg in the right eye and 18 mmHg in the left, and had no complaints. She was continued on Neptazane. On December 17, exactly 3 months after beginning Neptazane, the patient presented to her internist and was admitted to the hospital with a 1-week history of malaise, fever, a 4-pound weight loss and a "rash," which was ecchymosis. She was on no other medications besides Neptazane and had no history of sulfonamide allergy. Her leukocyte count was 1000 ~1 with 14 polymorphonucleocytes, 82 lymphocytes, and 2 monocytes; her platelet count was 20,000. Her
hemoglobin level was 8.3 g/dl. Two days after her admission, she died of sepsis, and Neptazane was suspected as the cause of her aplastic anemia. No autopsy was performed.
SUBJECTS AND METHODS Questionnaires were sent to all ophthalmologists listed in the 1983 to 1984 Directory of the American Academy of Ophthalmology in the Buffalo, Denver, and New Orleans areas, to obtain a geographic sampling of the country. They were surveyed as to their use of Diamox and Neptazane, their monitoring of patients on these drugs, and their assessment of the standard of practice in this regard. Of 128 ophthalmologists surveyed between January and March 1985, 81 who prescribe these drugs responded. Their responses are tallied. Fifty-eight members of Departments of Ophthalmology in 17 academic institutions around the country also were polled, and responses were received from 40, including 10 glaucoma specialists (Table 1). Five previous authors and one commentator1- 6 were contacted for information regarding their current practices and recommendations regarding blood monitoring. A separate survey was conducted among the participants in the 1985 Annual North American Glaucomatologists' Learning Ensemble regarding their blood count monitoring of patients on carbonic anhydrase inhibitors and the incidence of blood dyscrasias among their patients. Responses were received from all 66 glaucoma specialists in attendance and are tallied (Table 2). Three hematologists were consulted regarding the assignment of etiologies in cases of aplastic anemia and the mechanism of carbonic anhydrase inhibitor-induced dyscrasias. Le769
OPHTHALMOLOGY
•
JUNE 1988 •
derle Laboratories, the manufacturer of Diamox and Neptazane, provided information regarding the magnitude of the use of these drugs nationally and the mechanism of the side effects of concern.
RESULTS Among the 81 practitioners responding, representing 60% of those polled, a total of 11 monitor blood counts at varying intervals of patients on Diamox or Neptazane, or both. Two of 15 respondents from Buffalo monitor counts: one who prescribes only short-term Diamox checks initially, and the other who uses both Diamox and Neptazane checks every 6 months. Six of the 45 respondents from Denver monitor counts as follows: one initially and at 1 week, one initially and at 3 months, one at 1 month and 6 months, one who is semiretired and has only two patients on these medications monitors every 6 months, one annually, and one who uses the drugs only short term monitors blood counts if treatment lasts more than 10 to 14 days. Of 22 respondents from New Orleans, three monitor blood counts: one initially and at 6 months, one who uses only Diamox checks "occasionally," and one monitors only patients on Neptazane every 6 months. In assessing the standard of practice in their communities, four of these practitioners stated they did not know, and seven stated it was to do no testing. In total, 13.6% of practitioners monitor blood counts for one or the other drug or both at widely varying intervals. Sixtyfive percent believe the standard of practice to be no testing, 11% thought that they did not know the standard, and most respondents who named any following laboratory study as standard cited potassium rather than blood counts. Of 40 academic respondents, only two routinely monitor blood counts of their patients on carbonic anhydrase inhibitors, and two others check under certain conditions (Table 1). None of the 10 glaucoma specialists among this group checks blood counts routinely, and several commented emphatically that blood monitoring does not protect patients and is wasteful. Two of the six reporters, including one commentator who previously recommended routine blood counts, currently continue to recommend and practice such monitoring. One of these, 3 a hematologist who does not prescribe these drugs himself, continues to recommend monthly counts (personal communication). Another, 4 who ordered blood counts bimonthly for many years with negative results, currently monitors only patients on long-term Diamox at approximately 4-month intervals (personal communication, A. Johnston). In 1980, Werblin and coauthors, 1 referring to their cases that did not develop acutely, stated, "It is possible in these instances that periodic blood analysis may have detected the problem in time to prevent the severe reactions that later developed." 1 In 1985, they noted that patients on these drugs are seen 3 to 6 weeks after initiation oftherapy for reevaluation of symptomatology, but blood studies are not done routinely (personal communication). The remaining three authors who previously rec770
VOLUME 95
•
NUMBER 6
Table 2. Intervals of and Reasons for Blood Count Monitoring of Patients on Carbonic Anhydrase Inhibitors by 12 Glaucoma Specialists among 66 Responding Reasons for Testing Intervals of Testing Initially only Initially and 1 mo Initially and 3-6 mos Initially and 4 mos Initially and 12 mos 3-6 mo
Medical Only
Legal Only
Both
Not Specified
3
ommended routine counts2 •5•6 currently do not routinely order these (personal communications). Twelve of 66 glaucoma specialists who responded monitor blood counts of their patients on carbonic anhydrase inhibitors at varying intervals and for varying reasons (Table 2). None of these has ever had a blood dyscrasia develop in a patient on a carbonic anhydrase inhibitor. Six of the 66 glaucoma specialists reported seven blood dyscrasias developing in their patients who were on these drugs, again at varying intervals. One reported a fatal aplastic anemia at 2 months from initiation of the drug, and the other six cases were reversible marrow suppressions occurring at 2 months, 9 months (2 patients), 12 months (2 patients), and 3 years, respectively, from initiation of the carbonic anhydrase inhibitors. Interestingly, none of these specialists has monitored blood counts, even since their patients' blood dyscrasias. As in the other survey, several respondents decried the published recommendations for blood count monitoring as ill-advised and an outright disservice. The hematologists consulted agreed that although the mechanism by which these drugs cause blood dyscrasia is not known for sure, it is thought to be idiosyncratic and non-dose-related. There is no test to prove an etiologic relationship between a drug and a blood dyscrasia, 7 and the time of onset of such dyscrasias and the speed of development are both variable and unpredictable (personal communication, J. Fitchen, University of Oregon). Lederle Laboratories, the manufacturer of Diamox and Neptazane, report that a non-dose-related mechanism is postulated for blood dyscrasias associated with carbonic anhydrase inhibitors (personal communication, R. Caspari, Lederle), but their package inserts (Neptazane #85398015 and Diamox #15889013) recommend blood monitoring. The company was unable to more precisely define their terms "periodic" and "at regular intervals" with regard to monitoring, and in fact could not provide a rationale for suggesting routine monitoring (personal communication, N. Bishop, Lederle).
DISCUSSION In our industrialized society, nearly everyone is exposed to potentially toxic chemicals. In order for a single drug to be declared the probable cause of a blood dys-
MOGK AND CYRLIN
•
BLOOD DYSCRASIAS AND CAis
crasia (and this is the most that can be stated), an exhaustive history must be undertaken to exclude all other possible causative agents. This was not done in our first case, and in fact the date of starting Neptazane was mistated by 3 months. Similar deficiencies are seen in other reported cases. The National Registry of Drug-induced Ocular Side Effects itself has bemoaned the incomplete nature of the drug histories provided in some cases reported to them (personal communication, S. Meyer). If care is not taken, any anemia in a patient who happens to be on a carbonic anhydrase inhibitor will be blamed on the drugs. Carbonic anhydrase inhibitors are unsubstituted sulfonamides and manifest similar toxic and allergic properties, notably a variety of blood dyscrasias. The important features of these reactions are that the time between initiation of drug use and onset of blood dyscrasia is variable and although the mechanism is believed to be idiosyncratic, the majority of blood dyscrasias has been reversible upon discontinuation of the drug when symptoms occurred. Sixty-two of the 79 cases of possible carbonic anhydrase inhibitor-related blood dyscrasias reported since 1955 show times of onset ranging from less than 1 month to 13 months after beginning the drug. The majority of blood dyscrasias, 27 of 40 on acetazolamide and 17 of 21 on methazolamide, occurred within the first 4 months of drug treatment. 8 Outcomes of 74 cases include 26 fatalities, 35 recoveries, and 13 in the process of recovering at the time of reporting. 8 Are routine blood counts the solution to protection for those patients whose blood dyscrasias are responsibly thought to be secondary to carbonic anhydrase ingestion? The National Registry of Drug-induced Ocular Side Effects has urged blood counts at the initiation of treatment and at 6-month intervals, despite the fact that by their own data the large majority of dyscrasias occurs before 6 months. 8 The patients reported here had died well before 6 months from the time of initiation of the drug. We have seen that the vast majority of academicians, including glaucoma specialists who are in a position to use these drugs regularly, do not themselves monitor blood counts, nor do most practitioners, and several authors who have recommended it in the past no longer do. Does cost play a part in the decision? Lederle's market research has identified the number of new patients started on Diamox in 1984 as 160,000 and Neptazane 52,000 (personal communication, J. Briggs, Lederle). During that time, there was one fatal aplastic anemia in a patient taking Diamox for 71f2 months reported to the National Registry and one nonfatal case of hemolytic anemia and idiopathic thrombocytopenia purpura reported to Lederle. In addition, there are the two cases reported here of fatal aplastic anemia in patients taking Neptazane. If a baseline and one repeat test per new patient, at an expenditure of $8,480,000 (at $20 per blood count), would have saved three lives, one might certainly decide it was .worth doing. But it would not have saved these lives. If the tests had been done at 6 months, two of the three patients would have died be-
fore testing, and there is no assurance that the third patient would have shown any blood abnormalities as early as 6 months. In addition, there is no assurance that any of the blood abnormalities would have been reversible if detected at any time. The real problem with routine blood monitoring of patients on carbonic anhydrase inhibitors for early detection of blood dyscrasias is twofold. First, since the mechanism is not known but believed by most to be idiosyncratic and non-dose-related, discontinuation of the drug does not necessarily result in reversal of the blood dyscrasia no matter when it is discovered, although many discovered by symptoms have reversed. Second, since there is no pattern to the time of onset of the blood dyscrasias or to the speed of development from onset to complete aplasia, the term early as in early detection is undefinable. There is no optimum time to draw a complete blood count.
CONCLUSION Care should be taken when assigning an etiology to an aplastic anemia or other blood dyscrasia. The temptation to offer routine blood counts as an easy solution to patient protection should be resisted. Observation and questioning of patients on carbonic anhydrase inhibitors for symptoms of anemia, infection, or poor clotting is preferable, especially during the first 6 months of treatment. Since patients often report systemic symptoms to their internists rather than their ophthalmologists, there is need for communication between the two. Blood tests should be reserved for those patients whose complaints arouse suspicion.
ACKNOWLEDGMENTS The authors thank Patrick M. Verb, MD, for providing case
l.
REFERENCES 1. Werblin TP, Pollack IP, Liss RA. Blood dyscrasias in patients using methazolamide (Neptazane) for glaucoma. Ophthalmology 1980; 87:350-3. 2. Ellis PP. Discussion. Ophthalmology 1980; 87:350-3. Of: Werblin TP, Pollack IP, Liss RA. Blood dyscrasias in patients using methazolamide (Neptazane) for glaucoma. Ophthalmology 1980; 87:354. 3. Wisch N, Fischbein Fl, Siegel R, et al. Aplastic anemia resulting from the use of carbonic anhydrase inhibitors. Am J Ophthalmol 1973; 75:130-2. 4. Rentiers PK, Johnston AC, Buskard N. Severe aplastic anemia as a complication of acetazolamide therapy. Can J Ophthalmol 1970; 5:337-42. 5. Lubeck MJ. Aplastic anemia following acetazolamide therapy. Am J Ophthalmol1970; 69:684-5. 6. Turtz CA, Turtz AI. Toxicity due to acetazolamide (Diamox). Arch Ophthalmol1958; 60:130-1. 7. Erslev AJ. Drug-induced blood dyscrasias, 1. Aplastic Anemia. JAMA 1964; 188:531-2. 8. Fraunfelder FT. Meyer SM. Bagby GC Jr, Dreis MW. Hematologic reactions to carbonic anhydrase inhibitors. Am J Ophthalmol 1985; 100:79-81.
771
Abstract: Two new cases of aplastic anemia possibly associated with Neptazane (methazolamide) are reported. Several previous authors, as well as the manufacturer of Diamox (acetazolamide) and Neptazane, have recommended routine blood counts for patients on carbonic anhydrase inhibitors. Four surveys are presented of current practices with regard to blood monitoring. These include authors of case reports, 40 academic ophthalmologists, 81 ophthalmologists in private practice, and 66 glaucoma specialists. The vast majority does not routinely monitor blood counts of patients on carbonic anhydrase inhibitors. The value of routine blood monitoring is questionable both because of (1) the idiosyncratic, non-dose-related mechanism of the dyscrasias and (2) the variability of the timing of their onset and development. Regular observation and questioning of patients for symptoms are thought to be preferable. The importance of a thorough history when assigning an etiology to a dyscrasia is noted. [Key words: acetazolamide, aplastic anemia, blood dyscrasias, carbonic anhydrase inhibitor, Diamox, methazolamide, Neptazane.] Ophthalmology 95:768771' 1988
As the number of case reports of blood dyscrasias possibly associated with carbonic anhydrase inhibitors increases, the response of the authors has often been to encourage routine blood counts for early detection. Two cases presented herein beg the question as to whether such blood counts could have detected the dyscrasias and helped prevent the fatal outcome. To answer this question, we reviewed the current knowledge of the mechanism and pattern of carbonic anhydrase inhibitor-induced dyscrasias; we surveyed monitoring practices of ophthalmologists, previous authors, ophthalmologic academicians, and glaucoma specialists. In this article we present their practices and recommendations for monitoring patients on carbonic anhydrase inhibitors.
CASE REPORTS Case 1. A 71-year-old white man was first seen on May 24, 1984. He was using Carbachol twice daily and had visual Originally received: July 13, 1987. Revision accepted: January 7, 1988. 1
2
Kresge Eye Institute and Grosse Pointe Ophthalmology PC, Detroit. Franklin Eye Consultants, Detroit.
Reprint requests to Lylas Good Mogk, MD, Grosse Pointe Ophthalmology PC, 15401 E. Jefferson Ave, Grosse Pointe Park, Ml 48230.
768
acuity of 20/60 in the right eye and 20/30 in the left, intraocular pressures (lOPs) of 33 and 24 mmHg, and cup: disc ratios of 0.9 and 0.7. Carbachol was increased to three times daily and Timoptic 0.5% and Propine® 0.1% were added. On July 7, his visual acuity was 20/100 in the right eye (lOP, 23 mmHg) and 20/40 in the left (lOP, 21 mmHg), and he was started on Neptazane 50 mg twice daily. He was seen monthly. On September 21, with lOPs of 24 mmHg in the right eye and 19 mmHg in the left, an argon laser trabeculoplasty was performed on the right eye. On October 6, his lOPs were 23 mmHg in the right eye and 18 mmHg in the left. On October 12, the patient phoned to report weakness in his legs. He was instructed to discontinue Neptazane and come in. He did discontinue Neptazane on October 12 but did not return to the ophthalmologist. On October l 7, he was admitted by his internist to a local hospital with fatigue, chills, sore throat, lower extremity ecchymosis, and a history of a 13-lb weight loss over the preceding 3 months. The patient was not on medication, had no history of sulfonamide allergy, and no history of chemical or environmental exposures was elicited. Laboratory results were as follows: hemoglobin 5.5 g/dl; leukocyte count 1500 Ill with 99% lymphocytes; and platelets, 2000. Bone marrow biopsy done shortly after his October 17 admission showed acellular damaged marrow with fatty replacement, and the diagnosis of aplastic anemia was made. Throat cultures grew Staphylococcus aureus. Over the ensuing 2 weeks, the patient was treated with multiple antibiotics including Tobramycin, Ticarcillin, Vancomycin, amphotericin B, Cefabid, and Amikacin. Multiple transfusions of erythrocytes and platelets were given, the hemoglobin level rose to 8.6 g/dl, leukocyte count to 14,000 Ill, and platelet count to
MOGK AND CYRLIN
•
BLOOD DYSCRASIAS AND CAis
Table 1. Academicians Polled Regarding Blood Count Monitoring of Patients on Carbonic Anhydrase Inhibitors Institution Bascom Palmer Beth Israel, Boston Kresge Eye Institute Louisiana State University Massachusetts Eye and Ear Mt. Sinai, New York New York-Cornell New York University Ochsner Clinic Scheie Institute Sinai Hospital, Detroit Tufts New England Tulane University University of California at San Francisco University of Colorado Wills Eye Institute Wilmer Institute Total *
No. Contacted
No. Responding
No. Who Monitor
4* 1 2* 1 2* 2* 2* 1 3* 3 2* 3 3 2* 1 3 3*
0 0 0 1 1t 0 0 0 0 0 0 0 1 0 0 1:j: 0
5
1 2
5
4 2 2 3
5
4 2 4
5
2 1
5
4
40
56
Time Interval
Before treatment
Initially and every 3 mos Every 3 mos
4
Includes glaucoma specialist(s).
t Only if symptoms, not routinely.
:j: To monitor disease processes unrelated to carbonic anhydrase inhibitors.
59,000. On November 7, the patient's temperature was 105"F, and on November 9 he died of septic shock. No autopsy was performed. The death certificate cited methazolamide as the cause of the aplastic anemia. Case 2. A 73-year-old white woman was first seen on May 10, 1984, with a visual acuity of 20/30 in the right eye and 20/15 in the left, lOPs of24 mmHg in both eyes, and enlarged optic cups. She was on Timoptic 0.5% twice daily and on no other medications. Propine 0.1% was added, and 1 month later her lOPs were 23 mmHg in both eyes. Propine was replaced by Pilocarpine 1%, then 2%, and finally 4%. On September 17, with a visual acuity of 20/40 in the right eye and 20/25 in the left and lOPs of23 mmHg in both eyes, the patient was started on Neptazane 50 mg twice daily, and Pilocarpine was discontinued. On October 10, lOPs were 21 mmHg in the right eye and 20 mmHg in the left, and Neptazane was increased to three times daily. Argon laser trabeculoplasties were performed on October 15 and 29, and on October 31 lOPs were 16 mmHg in both eyes, and the patient had no complaints. She was continued on Neptazane. On November 13, the patient complained to her internist of diarrhea, for which she was taking Kaopectate. She was given lmodium, and a complete blood count was ordered which showed a leukocyte count of 3600 ~1 with 39 polymorphonucleocytes, 55 lymphocytes, 6 monocytes, and no bands; her platelet count was 257,000. The decreased leukocyte count did not arouse suspicion. The patient was seen by her ophthalmologist on December 5, with lOPs of 22 mmHg in the right eye and 18 mmHg in the left, and had no complaints. She was continued on Neptazane. On December 17, exactly 3 months after beginning Neptazane, the patient presented to her internist and was admitted to the hospital with a 1-week history of malaise, fever, a 4-pound weight loss and a "rash," which was ecchymosis. She was on no other medications besides Neptazane and had no history of sulfonamide allergy. Her leukocyte count was 1000 ~1 with 14 polymorphonucleocytes, 82 lymphocytes, and 2 monocytes; her platelet count was 20,000. Her
hemoglobin level was 8.3 g/dl. Two days after her admission, she died of sepsis, and Neptazane was suspected as the cause of her aplastic anemia. No autopsy was performed.
SUBJECTS AND METHODS Questionnaires were sent to all ophthalmologists listed in the 1983 to 1984 Directory of the American Academy of Ophthalmology in the Buffalo, Denver, and New Orleans areas, to obtain a geographic sampling of the country. They were surveyed as to their use of Diamox and Neptazane, their monitoring of patients on these drugs, and their assessment of the standard of practice in this regard. Of 128 ophthalmologists surveyed between January and March 1985, 81 who prescribe these drugs responded. Their responses are tallied. Fifty-eight members of Departments of Ophthalmology in 17 academic institutions around the country also were polled, and responses were received from 40, including 10 glaucoma specialists (Table 1). Five previous authors and one commentator1- 6 were contacted for information regarding their current practices and recommendations regarding blood monitoring. A separate survey was conducted among the participants in the 1985 Annual North American Glaucomatologists' Learning Ensemble regarding their blood count monitoring of patients on carbonic anhydrase inhibitors and the incidence of blood dyscrasias among their patients. Responses were received from all 66 glaucoma specialists in attendance and are tallied (Table 2). Three hematologists were consulted regarding the assignment of etiologies in cases of aplastic anemia and the mechanism of carbonic anhydrase inhibitor-induced dyscrasias. Le769
OPHTHALMOLOGY
•
JUNE 1988 •
derle Laboratories, the manufacturer of Diamox and Neptazane, provided information regarding the magnitude of the use of these drugs nationally and the mechanism of the side effects of concern.
RESULTS Among the 81 practitioners responding, representing 60% of those polled, a total of 11 monitor blood counts at varying intervals of patients on Diamox or Neptazane, or both. Two of 15 respondents from Buffalo monitor counts: one who prescribes only short-term Diamox checks initially, and the other who uses both Diamox and Neptazane checks every 6 months. Six of the 45 respondents from Denver monitor counts as follows: one initially and at 1 week, one initially and at 3 months, one at 1 month and 6 months, one who is semiretired and has only two patients on these medications monitors every 6 months, one annually, and one who uses the drugs only short term monitors blood counts if treatment lasts more than 10 to 14 days. Of 22 respondents from New Orleans, three monitor blood counts: one initially and at 6 months, one who uses only Diamox checks "occasionally," and one monitors only patients on Neptazane every 6 months. In assessing the standard of practice in their communities, four of these practitioners stated they did not know, and seven stated it was to do no testing. In total, 13.6% of practitioners monitor blood counts for one or the other drug or both at widely varying intervals. Sixtyfive percent believe the standard of practice to be no testing, 11% thought that they did not know the standard, and most respondents who named any following laboratory study as standard cited potassium rather than blood counts. Of 40 academic respondents, only two routinely monitor blood counts of their patients on carbonic anhydrase inhibitors, and two others check under certain conditions (Table 1). None of the 10 glaucoma specialists among this group checks blood counts routinely, and several commented emphatically that blood monitoring does not protect patients and is wasteful. Two of the six reporters, including one commentator who previously recommended routine blood counts, currently continue to recommend and practice such monitoring. One of these, 3 a hematologist who does not prescribe these drugs himself, continues to recommend monthly counts (personal communication). Another, 4 who ordered blood counts bimonthly for many years with negative results, currently monitors only patients on long-term Diamox at approximately 4-month intervals (personal communication, A. Johnston). In 1980, Werblin and coauthors, 1 referring to their cases that did not develop acutely, stated, "It is possible in these instances that periodic blood analysis may have detected the problem in time to prevent the severe reactions that later developed." 1 In 1985, they noted that patients on these drugs are seen 3 to 6 weeks after initiation oftherapy for reevaluation of symptomatology, but blood studies are not done routinely (personal communication). The remaining three authors who previously rec770
VOLUME 95
•
NUMBER 6
Table 2. Intervals of and Reasons for Blood Count Monitoring of Patients on Carbonic Anhydrase Inhibitors by 12 Glaucoma Specialists among 66 Responding Reasons for Testing Intervals of Testing Initially only Initially and 1 mo Initially and 3-6 mos Initially and 4 mos Initially and 12 mos 3-6 mo
Medical Only
Legal Only
Both
Not Specified
3
ommended routine counts2 •5•6 currently do not routinely order these (personal communications). Twelve of 66 glaucoma specialists who responded monitor blood counts of their patients on carbonic anhydrase inhibitors at varying intervals and for varying reasons (Table 2). None of these has ever had a blood dyscrasia develop in a patient on a carbonic anhydrase inhibitor. Six of the 66 glaucoma specialists reported seven blood dyscrasias developing in their patients who were on these drugs, again at varying intervals. One reported a fatal aplastic anemia at 2 months from initiation of the drug, and the other six cases were reversible marrow suppressions occurring at 2 months, 9 months (2 patients), 12 months (2 patients), and 3 years, respectively, from initiation of the carbonic anhydrase inhibitors. Interestingly, none of these specialists has monitored blood counts, even since their patients' blood dyscrasias. As in the other survey, several respondents decried the published recommendations for blood count monitoring as ill-advised and an outright disservice. The hematologists consulted agreed that although the mechanism by which these drugs cause blood dyscrasia is not known for sure, it is thought to be idiosyncratic and non-dose-related. There is no test to prove an etiologic relationship between a drug and a blood dyscrasia, 7 and the time of onset of such dyscrasias and the speed of development are both variable and unpredictable (personal communication, J. Fitchen, University of Oregon). Lederle Laboratories, the manufacturer of Diamox and Neptazane, report that a non-dose-related mechanism is postulated for blood dyscrasias associated with carbonic anhydrase inhibitors (personal communication, R. Caspari, Lederle), but their package inserts (Neptazane #85398015 and Diamox #15889013) recommend blood monitoring. The company was unable to more precisely define their terms "periodic" and "at regular intervals" with regard to monitoring, and in fact could not provide a rationale for suggesting routine monitoring (personal communication, N. Bishop, Lederle).
DISCUSSION In our industrialized society, nearly everyone is exposed to potentially toxic chemicals. In order for a single drug to be declared the probable cause of a blood dys-
MOGK AND CYRLIN
•
BLOOD DYSCRASIAS AND CAis
crasia (and this is the most that can be stated), an exhaustive history must be undertaken to exclude all other possible causative agents. This was not done in our first case, and in fact the date of starting Neptazane was mistated by 3 months. Similar deficiencies are seen in other reported cases. The National Registry of Drug-induced Ocular Side Effects itself has bemoaned the incomplete nature of the drug histories provided in some cases reported to them (personal communication, S. Meyer). If care is not taken, any anemia in a patient who happens to be on a carbonic anhydrase inhibitor will be blamed on the drugs. Carbonic anhydrase inhibitors are unsubstituted sulfonamides and manifest similar toxic and allergic properties, notably a variety of blood dyscrasias. The important features of these reactions are that the time between initiation of drug use and onset of blood dyscrasia is variable and although the mechanism is believed to be idiosyncratic, the majority of blood dyscrasias has been reversible upon discontinuation of the drug when symptoms occurred. Sixty-two of the 79 cases of possible carbonic anhydrase inhibitor-related blood dyscrasias reported since 1955 show times of onset ranging from less than 1 month to 13 months after beginning the drug. The majority of blood dyscrasias, 27 of 40 on acetazolamide and 17 of 21 on methazolamide, occurred within the first 4 months of drug treatment. 8 Outcomes of 74 cases include 26 fatalities, 35 recoveries, and 13 in the process of recovering at the time of reporting. 8 Are routine blood counts the solution to protection for those patients whose blood dyscrasias are responsibly thought to be secondary to carbonic anhydrase ingestion? The National Registry of Drug-induced Ocular Side Effects has urged blood counts at the initiation of treatment and at 6-month intervals, despite the fact that by their own data the large majority of dyscrasias occurs before 6 months. 8 The patients reported here had died well before 6 months from the time of initiation of the drug. We have seen that the vast majority of academicians, including glaucoma specialists who are in a position to use these drugs regularly, do not themselves monitor blood counts, nor do most practitioners, and several authors who have recommended it in the past no longer do. Does cost play a part in the decision? Lederle's market research has identified the number of new patients started on Diamox in 1984 as 160,000 and Neptazane 52,000 (personal communication, J. Briggs, Lederle). During that time, there was one fatal aplastic anemia in a patient taking Diamox for 71f2 months reported to the National Registry and one nonfatal case of hemolytic anemia and idiopathic thrombocytopenia purpura reported to Lederle. In addition, there are the two cases reported here of fatal aplastic anemia in patients taking Neptazane. If a baseline and one repeat test per new patient, at an expenditure of $8,480,000 (at $20 per blood count), would have saved three lives, one might certainly decide it was .worth doing. But it would not have saved these lives. If the tests had been done at 6 months, two of the three patients would have died be-
fore testing, and there is no assurance that the third patient would have shown any blood abnormalities as early as 6 months. In addition, there is no assurance that any of the blood abnormalities would have been reversible if detected at any time. The real problem with routine blood monitoring of patients on carbonic anhydrase inhibitors for early detection of blood dyscrasias is twofold. First, since the mechanism is not known but believed by most to be idiosyncratic and non-dose-related, discontinuation of the drug does not necessarily result in reversal of the blood dyscrasia no matter when it is discovered, although many discovered by symptoms have reversed. Second, since there is no pattern to the time of onset of the blood dyscrasias or to the speed of development from onset to complete aplasia, the term early as in early detection is undefinable. There is no optimum time to draw a complete blood count.
CONCLUSION Care should be taken when assigning an etiology to an aplastic anemia or other blood dyscrasia. The temptation to offer routine blood counts as an easy solution to patient protection should be resisted. Observation and questioning of patients on carbonic anhydrase inhibitors for symptoms of anemia, infection, or poor clotting is preferable, especially during the first 6 months of treatment. Since patients often report systemic symptoms to their internists rather than their ophthalmologists, there is need for communication between the two. Blood tests should be reserved for those patients whose complaints arouse suspicion.
ACKNOWLEDGMENTS The authors thank Patrick M. Verb, MD, for providing case
l.
REFERENCES 1. Werblin TP, Pollack IP, Liss RA. Blood dyscrasias in patients using methazolamide (Neptazane) for glaucoma. Ophthalmology 1980; 87:350-3. 2. Ellis PP. Discussion. Ophthalmology 1980; 87:350-3. Of: Werblin TP, Pollack IP, Liss RA. Blood dyscrasias in patients using methazolamide (Neptazane) for glaucoma. Ophthalmology 1980; 87:354. 3. Wisch N, Fischbein Fl, Siegel R, et al. Aplastic anemia resulting from the use of carbonic anhydrase inhibitors. Am J Ophthalmol 1973; 75:130-2. 4. Rentiers PK, Johnston AC, Buskard N. Severe aplastic anemia as a complication of acetazolamide therapy. Can J Ophthalmol 1970; 5:337-42. 5. Lubeck MJ. Aplastic anemia following acetazolamide therapy. Am J Ophthalmol1970; 69:684-5. 6. Turtz CA, Turtz AI. Toxicity due to acetazolamide (Diamox). Arch Ophthalmol1958; 60:130-1. 7. Erslev AJ. Drug-induced blood dyscrasias, 1. Aplastic Anemia. JAMA 1964; 188:531-2. 8. Fraunfelder FT. Meyer SM. Bagby GC Jr, Dreis MW. Hematologic reactions to carbonic anhydrase inhibitors. Am J Ophthalmol 1985; 100:79-81.
771