Five excerpts from discussions on the ferritin issue with Dr. Jerome Sullivan:

1]   Achieving a zero ferritin is diagnostically useful. At a zero level of stored iron, we can be fairly confident that a state of iron overload is not present at the time of the determination. The only way to have any real confidence that stored iron is not contributing to a patholgic process is to remove it and see if the patient does better clinically.

Patients with hepatitis C or diabetes may have an iron overload in this sense without having the hemochromatosis genes or the classic histologic findings of liver iron overload. If you insist that iron overload exists only when stringent, traditional criteria are satisfied, you may not find a higher proportion of iron overload thus defined in the study population. Nonetheless removal of small iron burdens may result in marked clinical improvement.

Attached below is the abstract of an important paper on hepatitis C and phlebotomy. Even after all stored iron was removed and the liver function tests improved, removal of more iron resulted in more improvement even after ferritin bottomed out. I would argue that these patients had a clinically significant degree of iron overload even
though the amount of iron in question was microscopic in comparison with the amounts seen in traditional iron overload.

Studies showing that a certain percentage of patients with this or that illness have iron overload are deeply flawed. By only looking at massive iron overload these studies are blind to any important effects of lower stored iron levels.

Jerome L. Sullivan, MD, PhD

Serum aminotransferase levels as an indicator of the effectiveness of venesection for chronic hepatitis C. Hayashi H , Takikawa T , Nishimura N , Yano M. J Hepatol 1995 Mar;22(3):268-71

BACKGROUND/AIMS: Iron cytotoxicity may play an important role in chronic hepatitis C. The effects of venesection suggest that a slight iron overload contributes to hepatic injury in subjects infected with hepatitis C virus. A better indication of the efficacy of venesection was studied in patients with and without overt iron overloading. METHODS: All 40 patients had chronic hepatitis C but none had hemochromatosis of a known etiology. A serum ferritin level of 10 ng/ml or less was chosen as the treatment goal. A mean blood volume of 2400 +/- 1100 ml was removed during treatments lasting 5 +/- 3 months. RESULTS: Treatment significantly reduced the mean serum levels of alanine aminotransferase activity from 128 +/- 74 to 63 +/- 28 IU/l (p < 0.01). The baseline enzyme activity was highly correlated with reduction in activity after treatment (r = 0.94, p < 0.01), but the baseline levels of ferritin and histochemistry for iron showed poor correlations with the reduction in enzyme activity (r = 0.63 with p < 0.01 and r = 0.38 with p < 0.05, respectively). CONCLUSIONS: Thus, serum levels of aminotransferases were a more important indicator for venesection than conventional indices of iron overload, probably because cytotoxic iron includes some reactive iron species rather than stored iron alone.

2]  Of course most hematologists believe in having some storage iron on board. But where does this opinion come from? It largely comes from what they were taught, not from wrestling with the angel and coming up with an interpretation based only on facts. All sciences are susceptible to uncritical acceptance of the professor's biases, but medical science is probably the worst. The incentives favoring consensus are much stronger in medicine. From a medicolegal point of view, a physician cannot be successfully sued for a course of action if the group of specialists in the relevant area all agree that it is right, even if it is in fact wrong. Thus there is a powerful impulse toward groupthink.

I think it is universally agreed that moment-to-moment normal resting physiological functioning is not impaired by total iron depletion, or even mild iron deficiency. There are some data that maximal performance may be very slightly impaired. These data are not, however, compelling. Recall that marathon runners are often mildly iron
deficient, because of iron loss in urine and gut. How can they run 26 miles at a stretch if absence of ferritin is harmful? Many women remain iron depleted or even mildly anemic from iron deficiency for their reproductive decades and still manage to outlive their husbands.

Certainly, severe anemia may be quite hazardous depending on the degree of the anemia and perhaps more importantly on the state of the patient's health. But STORED iron is not needed to prevent anemia.
So . . . what is the real value of having stored iron on board? It will certainly not help in the face of massive acute hemorrhage. The bone marrow cannot keep up with massive bleeding, no matter how much iron is in storage. The only benefit I can think of is the prevention of anemia from chronic blood loss. And this "benefit" may be more of a hazard than a true benefit. In the classic presentation of colon carcinoma the patient develops iron deficiency anemia from chronic occult blood loss.

A large load of stored iron would only serve to delay the life saving sign of anemia. With a typical load of 1000 mg and a bleeding rate of 2 mL per day (= 1 mg iron per day), the patient would not develop anemia until more than 1000 days after the tumor starts bleeding. Time enough for the tumor to penetrate the bowel wall and metastasize to the liver and other organs. If the patient had maintained 0 mg of iron in storage, the first mL of blood loss would translate into a lowered red cell mass, i.e. anemia and life saving symptoms.
Recall also that in the past there have been many occasions when all the specialists were flat wrong about this or that for prolonged periods, and then actively resisted correction of their erroneous positions.
Jerome L. Sullivan, MD, PhD

3]  Having a ferritin or zero and "having some iron around" are not mutually exclusive! With a zero ferritin, assuming hemoglobin level is within normal limits, there is a great plenty of iron around.  Remember that iron metabolism is dynamic. Iron is constantly being shuttled around even if there is flat zero STORED iron present. My
impression is that, indeed, the whole system gets better at getting out the last traces of stored iron as a mild anemia develops. Animal studies on the antioxidant effects of iron depletion/iron deficiency suggest that the body's systems for preventing iron toxicity become more effective if there is no stored iron around. These studies suggest a strong antioxidant boost after removal of stored iron.
Jerome L. Sullivan, MD, PhD

4]   I don't feel that arbitrary low target values for ferritin are very valuable.  Their main use seems to be to keep ferritin above zero, as if the patient would suddenly collapse if zero ferritin was reached.  I would remind the list again that volunteer blood donors are not required to have ferritin measured.  Ever!  That means that, so far as the FDA and the Blood Bank community are concerned, a volunteer donor's serum ferritin can fall straight to zero and stay there as long as the donor continues to donate regularly.  This is not at all an impossibility.  Donors are allowed to donate every 8 weeks and they are not required to take iron.  Remember that most donors do not have HH genes and so are not "protected" from a zero ferritin by the fast iron absorption of hemochromatosis patients.  Thus they are at increased "risk" of having a zero ferritin.  Does this matter?  No!  Volunteer donors do fine, in fact recent studies suggest that they do better that non donors in terms of heart attacks.

 Donors are required to have the hemoglobin level checked.  As long as they pass the hemoglobin test they can donate even if their ferritin is unmeasurably low.  A normal hemoglobin assures us that the donor has enough iron for all normal needs even if the donor has a flat zero level of stored iron.  All this stuff about having one iron loaded organ and one iron deficient organ at the same time is just idle speculation.  The totality of evidence says that a normal hemoglobin level means you have enough iron for all normal functions. Period.

 For hemochromatosis patients, there has developed an attitude that serum ferritin should be kept above some lower limit.  In my view, this is nonsense.  Why protect the ferritin level of hemochromatosis patients and let normal blood donors' ferritin fall without limit?  We must keep the fundamental goal of therapy firmly in mind:  We need to keep the stored iron level DOWN, not up!  If the serum ferritin slams down to flat zero, the patient should be congratulated, not made to worry, so long as hemoglobin remains normal.  For many patients a slightly lower than normal hemoglobin is fine if they tolerate it and if this can be achieved without an excessively burdensome phlebotomy schedule.  These are matters that must be fine tuned by the patient and his/her doctor.

 The serum ferritin value should be used to tell you when to keep being bled, not to tell you when to stop being bled.  The hemoglobin is the guide for temporarily stopping, not the ferritin. 

Jerome L. Sullivan, MD, PhD

5]  Let me say first that this discussion about when to stop or go with phlebotomy at the low end is less important for hemochromatosis patients than the initial bulk de-ironing.  In other words, it is far, far more important to go from a ferritin of 800 down to 50, than it is to go from 50 down to 0.   

What is needed in the zero to 50 ferritin range is a practical reliable way to assess progress.  In my view, the hemoglobin level provides the best guide.  Whatever guide is followed it must be practical and cost effective.  Hemoglobin is cheaper and more reliable than ferritin. 

The problem with ferritin is its fuzziness.  A ferritin of 50 ng/mL may be seen in a patient with zero iron stores.  But, a ferritin of 50 ng/mL may also be seen in a patient with substantial residual iron stores.   

A drop in hemoglobin level after a series of phlebotomies is the best evidence that the iron stores have been cleaned out, regardless of the ferritin level.  Maintaining a mild reduction in hemoglobin level may be too burdensome for some patients.  How well they tolerate the phlebotomy program is also an important consideration. 

A ferritin of 0 in the presence of a normal hemoglobin level would be unusual.  A hemochromatosis patient with these distinctly improbable numbers would be well advised to repeat the lab work.  The phlebotomy history and other lab work should be carefully reviewed, including MCV and possibly liver function tests.  In such a case it might be good to measure transferrin receptor levels. Determination of the patient's iron status would be a judgment call.  Certainly it would be appropriate to wait a few weeks and then re-examine the numbers before making a decision on the future phlebotomy rate.

 I would advocate a creative, individualized approach to fine tuning the phlebotomy rate.  Some patients and their physicians might want to consider liver function test (eg AST or ALT) results when deciding on when to do another phlebotomy.  LFTs often decline during the course of phlebotomy.  Some patients may want to maintain minimum LFT levels with phlebotomy.  This would be fine so long as a close eye is kept on the hemoglobin level.  Serum transferrin receptor/ serum ferritin ratios may be useful from time to time.  Generally bells and whistles beyond the hemoglobin measurement are not needed and add extra expense. 

Nothing beats having the patient intimately involved in the decision making.  You should become the world's expert on your case.  Follow those numbers.  Figure out how they change with volume of blood lost, and with how you feel.  Teach your doctor.  (And teach your relatives!) 

Jerome L. Sullivan, MD, PhD