Discussion
In our small cohort, we found a much higher incidence of individuals with documented 6-MMPN levels above 10 000 pmol/8 x 108 RBC than expected based on previous reports18,19; over 50% of our patients had shunting of 6-MP with associated toxicities. By contrast, previously documented rates of symptomatic 6-MMPN derangements in individuals were 20-30% in gastroenterology literature 19,20. It should be noted that the 6-MP dosing used in inflammatory bowel disease treatment is lower than that used in oncology. This higher dosing may explain the increased incidence of shunting in our oncology population. Our experience suggests that 6-MP metabolite levels should be monitored during maintenance therapy, particularly in the setting of known 6-MP toxicities, to prevent chemotherapy holds and dose reductions. In addition, because it is standard protocol to escalate 6-MP dosing when a patient’s absolute neutrophil count (ANC) is not adequately suppressed, we have noted that increases in 6-MP doses further exacerbate shunting and its adverse effects.
Allopurinol has a well-documented safety profile in its use for gout21. A similar safety profile was documented in our cohort. There were no direct adverse events after initiation of allopurinol in our patients despite the risk for increased myelosuppression if dosing is not carefully determined. Previously, because 6-MP was only available in 50 mg tablets, it was difficult to make fine dose adjustments in order to better establish the therapeutic dose. Combined therapy with allopurinol amplifies the clinical effect of 6-MP dose changes and it can cause undesired myelosuppression if dosing is not carefully titrated. We, therefore, provide an algorithm for combination therapy with a simple protocol for determining correct dosing for patients that has worked well for our center (Figure 3). In addition, the FDA approval of a liquid formulation of mercaptopurine makes it easier to manage combination therapy. In those patients in our series who were treated with allopurinol there was no apparent decrement in event free or survival outcomes.
Our series demonstrates efficacy for the use of allopurinol in the management of mercaptopurine side effects. Adverse effects related to mercaptopurine therapy including hepatotoxicity, hypoglycemia, and acute pancreatitis were reversed in all patients. While the chronic pancreatitis with insulin-dependent diabetes experienced by one patient was unchanged by the addition of allopurinol to the treatment regimen, this likely reflects the damage from the initial insult to the pancreas and not a lack of improvement with allopurinol. Patients started on combination therapy with allopurinol experienced reversal of undesired toxicities, suggesting that combination therapy may be beneficial for many more patients in the future.
Interestingly, a much higher percentage of males than females experienced shunting in our cohort. One possible explanation for this difference is the higher levels of TPMT activity often found in males22 that could theoretically result in higher levels of 6-MMPN levels if this higher activity were not balanced by the other enzymatic pathways. TPMT activity has also been shown to have racial differences, with Caucasians having higher TPMT levels, which would suggest a higher rate of shunting among Caucasian patients23. Perhaps due to our small cohort size we did not find significant differences between ethnicities, although it is interesting to note that 4 of the 5 African American patients in our cohort were ‘shunters. Of course, larger studies may be able to examine this finding in more detail and are warranted.
It has become our institution’s practice to monitor for signs of aberrant mercaptopurine metabolism, including hepatotoxicity, neutropenia (or neutrophilia), hypoglycemia, and pancreatitis. In these settings we begin monitoring 6-MMPN and 6-TGN levels. For those with markedly elevated 6-MMPN >10 000 pmol/8 x 108 RBC, we consider the addition of allopurinol. Considerations for the addition of allopurinol in each case include severity of toxicity and symptoms as well as compliance with current chemotherapy regimens. For those patients who agree to treatment with the combination of allopurinol-mercaptopurine we start with an age-based oral allopurinol dose (50 mg daily for age <5 years, 100 mg daily for age 5-10 years, 200 mg daily for age 10-15 years, and 300 mg daily for those over 15 years). Because of the enzyme blockade by allopurinol, we concomitantly reduce 6-MP to 25mg/m2/dose. For our younger children, utilizing liquid mercaptopurine can be beneficial in order to have tighter titration of the dose and effect. We then adjust 6-MP and allopurinol doses to achieve a goal of normalization of toxicities, absolute neutrophil count in the therapeutic window of 500-1 500 cells/µL and a 6-MMPN level of <5 000 pmol/8 x 108 RBC. If the 6-MMPN level is not substantially decreased with the addition of allopurinol, a higher dose of allopurinol dose should be attempted as there are individual variations in allopurinol effectiveness that require some patients to be on a higher allopurinol dose. Steroid and vincristine dosing during maintenance were not adjusted. Methotrexate dosing was also not adjusted during early phases of allopurinol-mercaptopurine titration, however if after several months ANC levels persisted >1500 cells/µL, then increasing methotrexate to 125% dosing was considered. Figure 3 offers an algorithmic approach to management.
It should be noted that the mechanism by which the addition of allopurinol to mercaptopurine therapy results in increased 6-TGN and decreased 6-MMPN is not fully understood. Preliminary laboratory experiments performed at our institution showed that allopurinol may cause decreased 6-MMPN levels due to tissue specific effects of 6-MP metabolism24. When 6-TGN and 6-MMPN levels are measured clinically, the levels are determined from red blood cells which are not undergoing cell division. In previous treatment protocols incorporating 6-Thioguanine (6-TG), which is directly converted to 6-TGN without any other metabolism, the 6-TG was found to be more hepatotoxic than 6-MP. Patients who are clinically treated with combination therapy have much higher 6-TGN levels than patients who are only on 6-MP and thus previous guidelines for 6-TGN therapeutic levels cannot be used to determine dosage efficacy. In addition, the extremely high 6-TGN levels, we achieved, with combination therapy while known to cause myelosuppression, did not result in hepatotoxicity. Based on our preliminary data, we suggest these changes in metabolite levels are due to the tissue specific effects of allopurinol on 6-MP metabolism; but further studies are required to confirm this finding.
One major limitation in this analysis is missing data for some individuals. Metabolite data was not checked in all of those who were eligible. Additionally, the frequency of checking metabolite data in patients also differs among those who were started on allopurinol. Of those who were started on allopurinol, Patients 2 and 12 were started based on clinical toxicities and did not have metabolite levels checked prior to allopurinol start. It is important to note that this lack of metabolite data could potentially indicate that our findings are an underestimation of the incidence of mercaptopurine metabolite shunting. Additionally, 6-MMPN and 6-TGN levels for our patients who did not exhibit clinical toxicities were not obtained for most of the cohort; these levels could have been useful to better establish a normal profile of 6-MMPN and 6-TGN levels, as well as a therapeutic window for the ratio of 6-MMPN:6-TGN which could be useful for guiding the initiation and benefits of allopurinol to mercaptopurine therapy. With use of this proposed algorithm, we think combination therapy has the potential to become routine practice in oncology.