Acute myeloid leukemia

Jump to Section

Acute myeloid leukemia

What every physician needs to know:

Acute myeloid leukemia (AML) is a cancer of blood forming cells. Many principles of cancer therapy have been developed in the context of treating AML, making it a paradigm for cancer chemotherapy. Ready access to tumor tissue (the malignant cells circulating in blood) has allowed tumor cell analysis and has increased our understanding of cancer pathophysiology through the discovery of genetic abnormalities within myeloid leukemia cells.

AML occurs when normal blood cell formation, characterized by a series of maturational divisions, in which hematopoietic stem cells give rise to normal blood elements including red cells, platelets, and white cells, is deranged, such that proliferation is greatly favored at the expense of normal maturation. An accumulation of malignant hematopoietic cells termed leukemic cells or “blasts” fills the marrow space and often spills out into the peripheral blood. Many patients with leukemia present with high white blood cell counts (white count) in which the blood cells are replaced by these immature myeloid blasts.

The leukemic cell is characterized by a high nuclear-to-cytoplastic ratio and a very immature appearing chromatin pattern. The clinical problems stemming from this unbridled proliferation and accumulation of immature hematopoietic cells and concomitant bone marrow failure include anemia (producing fatigue and shortness of breath), thrombocytopenia (leading to bleeding especially in mucosal sites), and neutropenia/leukopenia (resulting in infections with commensal organisms, especially gut and skin flora). If the blast count in the blood is very high, the leukemic cells themselves can adhere and cause microthrombi/microhemorrhage leading to deteriorating mental status and/or respiratory failure (so-called "leukostasis").

The clinician must also be alert to potential metabolic problems engendered by the leukemic cells. In certain subtypes of AML, particularly those in which the malignant cells are monocytoid, renal losses of potassium and/or bicarbonate may occur. Due to the frequent cell turnover and release of intracellular contents of the leukemic cells, hyperphosphatemia (with associated hypocalcemia) and hyperuricemia may occur. With hyperuricemia, uric acid deposition in the renal tubules may lead to renal failure, thereby exacerbating hypercalcemia and hyperphosphatemia. Thus, AML can present with fatigue, infections, bleeding, and/or renal failure.

Such devastating acute presentations are sometimes seen, but AML may also present in a more chronic way with progressive fatigue, especially in the case of older adults. In fact, in contrast to what is typically seen in tertiary care settings, AML is a disease of older adults with a median age of approximately 68 years.

Are you sure your patient has acute myeloid leukemia? What should you expect to find?

The diagnosis of acute myeloid leukemia (AML) is fairly straightforward, although a patient may sometimes present in an asymptomatic fashion when a complete blood count check is ordered for other reasons or for "routine screening". Patients generally present with one of the signs or symptoms that is associated with bone marrow failure. Fatigue or dyspnea, due to anemia and cytokine release from the malignant myeloblasts are frequently noted; thrombocytopenia leads to abnormal bleeding, specifically in mucosal surfaces such as the nose and gums and/or fine punctate bleeding in the skin, noted as red dot-shaped macules particularly prominent on the ankles (petechiae).

Patients may present with an infection, particularly pneumonia or soft tissue infection, due to the neutropenia. However, symptoms can be highly non-specific. Occasionally, patients can present with extramedullary leukemia; leukemic cells may or may not be prominent in the bone marrow, but more so in non-marrow sites. Particularly in the monocytic subtypes of AML, there may be deposits of leukemic cells in the skin, meninges, gums or other sites.

Occasionally, a dentist will refer a patient because of gingival hypertrophy. Cranial neuropathies caused by basilar skull infiltration or headache and vomiting due to increased intracranial pressure from decreased cerebrospinal fluid flow can be manifestations of leukemic meningitis. Probably the most common extramedullary manifestation of AML is leukemia cutis (leukemic infiltration of the skin), generally manifesting as pearly or opalescent nodules.

By far the most important presenting manifestation of AML is an abnormal blood count. Patients generally have a normocytic anemia and thrombocytopenia. While generally elevated, the white count may be low, particularly in most patients with the subtype of AML called acute promyelocytic leukemia (APL) or in some older adults with a more smoldering type of leukemic presentation.

Of particular importance is the blood smear which usually shows the presence of blasts. Blasts are immature hematopoietic cells with a high nuclear to cytoplasmic ratio. About 50% of patients with AML will have blasts that display Auer rods, a concretion of peroxidase/lysozyme-containing granules which appear as rod-like structures singly or in clusters in the cytoplasm of myeloblasts, and are particularly prominent in APL.

Disseminated intravascular coagulopathy (DIC), the laboratory abnormalities of which include an elevated thromboplastin time and low fibrinogen is a hallmark of the presentation of APL, but may be seen in any subtype of AML. This unbridled activation of the clotting cascade is believed to be due to the release of tissue factor from leukemic blasts during cell turnover. Patients with DIC and thrombocytopenia, as is typically seen in APL, may present with devastating bleeding complications such as major intracranial hemorrhage, which can be a terminal event in some cases.

Beware of other conditions that can mimic acute myeloid leukemia:

Patients with abnormal blood counts can have a host of conditions other than AML, though an elevated white blood count with more than 20% peripheral blasts, anemia and thrombocytopenia is a typical presentation for acute leukemia and is rarely confused with other conditions.

Probably the most important condition that mimics AML is acute lymphoblastic leukemia (ALL). This second type of acute leukemia can present in a similar fashion to AML but the blasts are lymphoid rather than myeloid in origin. Lymphoid blasts tend to have less cytoplasm, fewer or no granules, as well as fewer nucleoli than in myeloblasts. Auer rods are not seen in ALL, but the absence of Auer rods, present in only 50% of AML, does not exclude acute myeloid leukemia.

In the absence of Auer rods on Wright-stained smears, another level of testing is required to differentiate between lymphoblasts and myeloblasts. Such tests historically involved cytochemical analysis, which would show that lymphoblasts stain for the PAS (periodic acid-Schiff) enzyme, whereas myeloblasts stain for the peroxidase and/or non-specific esterase enzymes. Currently, immunophenotypic analysis is the main technique used to delineate lymphoid antigens versus myeloid antigens; although rarely the case of leukemia may have an ambiguous or mixed lineage..

Other states in which the white count can be high include leukemoid reactions, chronic myeloid leukemia (CML), or lower grade indolent lymphoproliferative neoplasms such as chronic lymphocytic leukemia or hairy cell leukemia. The latter entities are generally easily distinguished because of the characteristic mature appearing lymphocytes. In chronic phase CML, or leukemoid reaction there should be few or no blasts and a preponderance of mid-range myeloid cells such as bands, myelocytes, metamyelocytes, and promyelocytes.

Two important other related entities to AML in which peripheral myeloblasts can be seen include myelodysplastic syndromes (MDS) and myeloproliferative neoplasms. MDS, formerly termed "pre-leukemia", can include patients who have up to 20% blasts in the marrow and/or the peripheral blood. Patients with MDS, particularly those with excess blasts (between 5 and 20% marrow blasts) frequently convert to AML after a period of months to years. The distinguishing features between AML and higher grade MDS are purely quantitative.

The myeloproliferative neoplasms, particularly myelofibrosis (MF), can present with elevated white counts, some degree of peripheral blasts, and abnormal platelet and white counts. However, bone marrow examination in the case of MF would show a relatively small number of myeloblasts in the context of infiltration with fibroblasts.

Which individuals are most at risk for developing acute myeloid leukemia:

The most important risk factor for AML is age, probably due to aging of the bone marrow-stem cell compartment including the accumulation of genetic lesions over time.

Most patients with AML have no predisposing factors, however, chemotherapy for other cancers, particularly alkylating agents or topoisomerase II inhibiting drugs can be leukemogenic. In the case of alkylating agents, patients usually have a 5 to 8 year latency period and often have a myelodysplastic prodrome. and characteristic chromosomal abnormalities (loss of all or the long arm of chromosomes 5 and or 7, or greater than 3 distinct abnormalities (complex). In the case of the topoisomerase inhibitors such as etoposide, patients tend to have a shorter latency period (2 to 3 years) and often present with a monocytoid type of AML which has a characteristic translocation involving the long-arm of chromosome 11 at the MLL gene locus. Sometimes other chromosomal abnormalities - even those typically seen in patients with de novo AML or APL - can be seen in patients who have been previously exposed to radiation and/or chemotherapy. Recent molecular data also suggest that not every patient with so-called therapy-related AML has genetic changes suggestive of a chronic marrow stem cell disorder.

Other risk factors for AML include exposure to ionizing radiation from military, industrial, or therapeutic sources and exposure to certain industrial solvents such as benzene.

What laboratory studies should you order to help make the diagnosis and how should you interpret the results?

The most important laboratory studies in a patient with known or suspected AML are bone marrow aspirate and biopsy.

One should first review routine laboratory studies including a chemistry panel and complete blood count (CBC). Particular attention should be paid to the serum potassium which can be elevated due to tumor lysis, or depressed due to myeloblast (usually monocytoid blasts)-induced renal tubular electrolyte losses. The serum phosphate and uric acid may be quite high in tumor lysis syndrome, with hyperphosphatemia leading to hypocalcemia.

There are many reasons why the hepatic transaminases could be elevated in AML, including infection or leukemic infiltration of the liver. In the latter case, the serum alkaline phosphatase are elevated. Direct hyperbilirubinemia may occur due to biliary stones or infections; rarely indirect hyperbilirubinemia can indicate tumor cell-mediated hemolysis due to antibody elaboration. A low serum albumin might indicate a more chronic presentation with evidence of malnutrition at the time of diagnosis. Moreover, anthracyclines, which are required for induction therapy in AML, are excreted hepatically; direct hyperbilirubinemia would require at least consideration for dose modification (unless the abnormal liver tests can be definitively shown to be due to leukemic infiltration).

Finally, an assessment of the serum creatinine and blood urine nitrogen should be made, to help assess the patient’s hydration state and whether or not renal failure might be ongoing. Renal failure could indicate underlying renal dysfunction, especially in older patients, or, if acute, could indicate renal damage from uric acid released in the context of tumor lysis syndrome.

The CBC is a critical test in patients with known or suspected leukemia. Profound anemia, leukocytosis, and thrombocytopenia are common, although leukopenia can sometimes occur particularly in older adults with smoldering leukemia or in a patient who presents with APL. It is very important to review the peripheral blood smear because there are often circulating tumor cells (myeloblasts), which can be noted on a standard Wright-Giemsa stain.

AML patients’ blasts will contain Auer rods about 50% of the time. Auer rods are concretions of peroxidase-containing granules. AML blasts tend to have more nucleoli, more open chromatin, and somewhat granulated and more abundant cytoplasm than would be seen in patients with ALL. The definitive diagnosis of AML on a blood smear would only be possible if more than 20% of the white cells were blasts and if at least some of these cells contained Auer rods; otherwise, examination of the bone marrow or peripheral blasts using cytochemical and/or immunophenotypic analysis would need to be undertaken.

Especially in patients with APL, but in all patients, an INR (international normalized ratio), partial thromboplastin time, and fibrinogen should be checked. Almost all patients with APL and a minority of patients with other types of AML will present with disseminated intravascular coagulopathy (DIC) due to the release of pro-coagulant granules from the malignant cells. In full-blown cases of DIC, the thromboplastin time will also be elevated. The degree of hypofibrinogenemia correlates with the severity of the DIC.

Another laboratory study that may be useful early in the patient’s course is HLA (human leukocyte antigen) typing. Firstly, this is useful in cases where platelet allo-immunization becomes a problem, and it is necessary to transfuse HLA-matched platelets. Secondly, the majority of AML patients should be considered to be stem cell-transplant candidates, at least at diagnosis.

A bone marrow biopsy and aspirate should be carried out in every patient with suspected AML. The WHO (World Health Organization) classification system indicates that patients who have greater than 20% myeloblasts in the bone marrow have AML. Myeloblasts can be distinguished by their characteristic appearance including prominent nucleoli, a few cytoplasmic granules, and open chromatin. Definitive diagnosis of a myeloblast requires the presence of Auer rods, or cytochemical or immunophenotypic studies revealing myeloid derivation.

Cytochemistry should be performed on the bone marrow aspirate smear including assays for peroxidase (positive in most subtypes of AML), non-specific esterase (positive in monocytic subtypes of AML), and periodic acid Schiff (PAS positive in a chunky or block distribution in 50% of patients with ALL). An aliquot of bone marrow aspirate should be sent for flow cytometric analysis to definitively diagnose AML by virtue of positive detection of myeloid antigens such as CD33 (present in 90% of patients with AML), CD34 (stem cell antigen), CD13, CD15, CD11, CD14, and CD117. The HLA-DR antigen is positive in most patients with AML, but notably absent in patients with APL. In blasts that are cytochemically negative, a diagnosis of M0 AML (according to the old FAB [French-American-British] classification system) may be made if myeloid antigens (with the absence of lymphoid antigens) are noted on immunophenotypic studies. Blasts occasionally display no lineage-specific markers, or markers indicative of both myeloid and lymphoid derivation (mixed-phenotype acute leukemia).

An aliquot of aspirate marrow should be sent for cytogenetic analysis. Cytogenetics remains a key test to establish prognosis, understand pathophysiology, and, in many cases, guide therapy. Two cytogenetic abnormalities, inversion of chromosome 16 and translocation t(8;21), are characteristic of so called core-binding factor leukemias which have a favorable prognosis. Patients who have the type of chromosomal abnormalities seen in secondary AML or MDS, including loss of the long-arm or all of chromosome 5, loss of the long-arm or all of chromosome 7, and complex karyotype (greater than three abnormal cytogenetic findings) have a very poor prognosis. Monosomal karyotype (two chromosome deletions or one chromosome deletion plus structural abnormalities) augers an inferior prognosis, even when compared to the average patient with adverse cytogenetics.

About 70% of patients with AML have intermediate cytogenetic findings, which include a normal karyotype, an extra copy of chromosome 8 (trisomy 8), or translocations involving the 11q23 locus where the MLL gene resides. Cytogenetics can be used to confirm the diagnosis of APL. In such cases where APL is strongly suspected and cytogenetics are not technically feasible or will take excessive time to obtain results, important alternative diagnostic studies include FISH (fluorescence in situ hybridization) and PCR (polymerase chain reaction). FISH can be used to detect the characteristic t(15;17) translocation of APL (variant translocations can occur). FISH technology does not require cells to go into cell division and can be done on non-dividing cells. Perhaps even more rapid than FISH analysis, PCR detection of the fusion gene product, namely the PML-RARa transcript elaborated by the t(15;17) translocation, is pathognomonic of APL.

Testing for certain genetic abnormalities at diagnosis is increasingly becoming the standard of care. Patients with normal karyotype can be parsed into different prognostic categories by mutations occurring in certain genes. This is an evolving field as new genes with prognostic impact are being rapidly discovered.

At present, it is very helpful to know if normal karyotype patients have an FLT3 (Fms-like tyrosine kinase 3) ITD (internal tandem duplication) and/or an NPM1 mutation. An FLT3 ITD results from a duplication of between three and over 300 base pairs of the juxtamembrane region of the FLT3 transmembrane tyrosine kinase; it occurs in about 25% of patients and portends for a poor prognosis.

About 5 to 10% of patients have a point mutation in the tyrosine kinase domain of this enzyme, usually in the 835th amino acid residue, the prognostic significance of which is unclear. Another 30% of patients (partially overlapping with those who have a FLT3 mutation) have a mutation in the NPM1 gene which encodes for a nuclear to cytoplasmic shuttle protein. A mutation of NPM1, in the absence of a FLT3 mutation portends a good prognosis.

About 25% of patients with inversion 16 or t(8;21) chromosomal abnormalities have a mutation in the tyrosine kinase c-Kit which identifies a poor prognostic subset in those with an otherwise favorable core-binding factor leukemia. Finally about 5% of patients have a point mutation in the CEBP-a transcription factor. If this mutation is present in both alleles this predicts a very good prognosis.

Thus, at this time, most patients should have a diagnostic aliquot of peripheral of bone marrow blasts sent for FLT3, NPM1, and CEBP-amutational analysis. The list of genes that will be prognostically useful is expanding; ongoing studies are evaluating the prognostic and pathophysiological importance of genes that may epigenetically affect expression of multiple genes including TET2, IDH1, IDH2, and DNMT3A. Moreover, such findings (especially in the case of IDH, FLT3, KIT, or RAS mutations) may guide the clinician to be ready to refer for a ‘targeted’ therapy in development should the patient relapse or not be candidate for standard chemotherapy.

What imaging studies (if any) will be helpful in making or excluding the diagnosis of acute myeloid leukemia?


If you decide the patient has acute myeloid leukemia, what therapies should you initiate immediately?


Although some patients with acute leukemia present in an indolent fashion, particularly older adults, the majority of patients with AML present with issues that must be dealt with urgently. If there are any obvious infectious manifestations at presentation, appropriate antibiotics should be started. Because of the neutropenia and the lack of inflammatory cells, it can sometimes be difficult to delineate the source of the infection. Broad spectrum antibiotics covering commensal organisms from the gastrointestinal (GI) tract and the skin should be initiated in case of neutropenia and fever.

Hydration and blood product support

The patient’s hydration status should be monitored; IV (intravenous) saline should be given in patients who are dehydrated due to renal losses, vomiting, or lack of oral intake. Patients who are bleeding and thrombocytopenic should be given platelet transfusions; prophylactic transfusions are indicated if the platelet count is less than 10,000. Patients with APL or other leukemias associated with DIC should be treated rapidly with appropriate chemotherapy (see "More definitive therapies" below), but as a temporizing measure, platelet transfusions and factor replacement therapy (FFP or cryoprecipitate) should also be administered.

Reducing serum uric acid

If the serum uric acid is elevated and/or if there are other signs of tumor lysis (hyperphosphatemia or hypocalcemia), alkalinization of the urine to prevent uric acid deposition should be considered but is rarely needed in AML. Efforts to lower the serum uric acid should be accomplished. In cases where rapid reduction of the serum uric acid is not required, prophylactic allopurinol, a xanthine oxidase inhibitor, should be administered.

Full-blown tumor lysis syndrome with ongoing or potential renal failure, or if the patient cannot take oral drugs, should prompt administration of recombinant urate oxidase. Most patients will have a rapid reduction in their serum uric acid with a single 6mg dose of recombinant urate oxidase, but the dose can be repeated in 24 hours if needed.

Lowering the peripheral blood blast count

The most important consideration for urgent management of patients with AML is whether or not efforts are required to lower the peripheral blood blast count to treat or prevent "leukostasis" complications. Some AML patients with high blast counts, by virtue of the "stickiness" of the malignant cells, will have aggregation of such cells in the cerebral or pulmonary microvasculature. These microthrombi, often accompanied by microhemorrhages, can lead to the clinical sequelae of altered mental status or respiratory failure.

In general, saline hydration and the use of the rapidly acting ribonucleotide reductase inhibitor hydroxyurea can lower the white count over a few days, while the diagnosis is being made and definitive antileukemic therapy (see "What if scenarios" below) is being initiated.

The dose of hydroxyurea used in these situations should be 30 to 40mg/per kg/per day in single or divided doses (twice daily). If hydroxyurea cannot be given or if the patient has ongoing leukostasis and particularly if rapid reduction is needed, then leukopheresis should be initiated. Large bore intravenous therapy is required for this procedure which can rapidly reduce the white count, albeit temporarily, and improve central nervous system or pulmonary function. In general, no more than 2 to 4 leukopheresis procedures are needed to stabilize the patient until definitive anti-leukemic therapy can be given.

Occasionally low-dose radiation therapy can be used urgently to lyse cerebral myeloblasts, in cases of mental status deterioration, when leukopheresis cannot be done, or is ineffective.

More definitive therapies?

The definitive therapeutic approach to AML depends on specific disease and patient-related features. It is helpful to divide the therapeutic approaches into those pertaining to younger adults (less than age 60 with non-AML APML [acute promyelocytic leukemia]), older adults (greater than age 60) with non-APL AML, and patients with APL of any age.

The treatment paradigm for AML is divided into two stages; the first is induction therapy, and the second is post-remission consolidating therapy. In the first stage, chemotherapy is given to reduce the tumor burden (which is believed to be about 1012 cells at diagnosis) down to a level low enough so that a complete remission is achieved. Complete remission is defined as a state where leukemic cells are not detected by routine means in the bone marrow, peripheral blood, or extramedullary sites. Theory suggests that remission occurs when there are still 109 cells present; detectability is not straightforward. The use of multiparemeter flow cytometry or molecular analysis (particularly the detection of the PML-RARA fusion transcript in APL by PCR) can detect sub-morphological amounts of AML cells at the time of remission. The presence of such minimal residual disease (MRD) may portend a poor long-term outcome.

If a patient does not receive additional therapy in the post-remission setting, relapse is inevitable. A major effort has been dedicated to determining which post-remission therapy option (high-dose chemotherapy alone, high-dose chemotherapy with autologous stem cell support, or allogeneic stem cell transplantation) is the optimal approach for a given patient.

Younger adults with non-APL AML

Moreover, younger patients with non-APL AML cannot be considered as a single entity since the current treatment algorithm indicates that post-remission therapy should involve a risk-adapted approach. In general, this means that those with higher risk disease (less likely to remain in long-term remission with a chemotherapy alone strategy) should be allocated to receive an allogeneic stem cell transplant. Even using aggressive induction and risk-adapted post-remission therapy, only about 40% of younger adults with AML will experience long-term disease free survival.

Induction therapy should consist of an anthracycline-like agent for 3 days, in conjunction with cytarabine for 7 days. While various anthracyclines or anthracycline-like agents have been compared to standard daunorubicin, little consensus has emerged as to the optimal agent. Some studies suggest that idarubicin might be better than daunorubicin, but questions of dose equivalence made such comparisons difficult. A very important prospective randomized trial showed that 3 days of treatment at a dose of 90mg/m2 of daunorubicin was superior in terms of overall survival, than was 45mg/m2 per day (for 3 days) for the vast majority of younger patients with AML. Patients who had an FLT3 ITD mutation, those with high white count, or those between ages 55 and 60 did not benefit from this more dose-intensive approach.

Nonetheless, it is reasonable to use daunorubicin at 90mg/m2 per day for 3 days, along with cytarabine at 100 to 200mg/m2 per day by continuous IV infusion for 7 days. Unfortunately, during the time that this study was conducted, most trials were using daunorubicin in 60mg/m2, so it remains unknown whether “90 is better than 60”. If this 3+7 strategy is used, a “mid-cycle” bone marrow examination should be done 15 or 22 days after the start of chemotherapy. If this bone marrow is adequately cellular (greater than 10% cells) and harbors more than 10% leukemic blasts, than a reinduction strategy either consisting of another "3+7" chemotherapy, or 2 days of daunorubicin plus 5 days of cytarabine should be administered. The overall complete remission rate for younger adults with AML is approximately 70%. An alternative induction including idarubicin alone and continuous infusion high-dose ara-C (cytarabine) is used at the MD Anderson Cancer Center.

There have been many attempts to improve upon the "3+7" regimen commonly used to treat patients with AML for induction therapy. Several randomized trials in both younger and older adults have attempted to determine if idarubicin or mitoxantrone might be superior to daunorubicin within the "3+7" regimen. Proof of superiority of one anthracycline or one anthracycline-like agent compared to daunorubicin has been lacking; doxorubicin is probably inferior based on a now "old" trial, which showed that doxorubicin was associated with more gastrointestinal side effects, without any benefit in terms of efficacy. One problem with comparison of one drug to another is the question of dose intensity.

The dose of cytarabine has also been investigated during induction. A trial conducted 30 years ago by the CALGB (Cancer and Leukemia Group B) showed that cytarabine 100mg/m2 per day for 7 days was equivalent to 7 days at a dose of 200mg/m2 per day. Another question has been whether or not intensifying cytarabine during induction might be beneficial. While a phase II trial with high-dose cytarabine at the conclusion of "3+7" (3+7+3) seemed to show a high complete remission rate, a prospective comparison of "3+7" to "3+7+3" did not show a benefit to the more intensive regimen. Replacing high-dose cytarabine with standard dose cytarabine has also been attempted, with results suggesting a possible benefit in disease free but not overall survival for those receiving a more intensive induction regimen.

Another induction strategy is the "timed-sequential" approach in which a second course of induction therapy is administered at a fixed time after the first course. Giving chemotherapy during leukemic stem cell recovery would presumptively be effective, since many cells would be going into S phase and be susceptible to chemotherapy. Although such a regimen in children seems to be highly effective, there are only uncontrolled reports of the efficacy of timed-sequential approach in adults. This method has not been widely adopted.

A variation of the timed-sequential approach is represented by administration of hematopoietic growth factors prior to chemotherapy, in an attempt to maximize the number of leukemic cells at S phase prior to the use of cytotoxic agents. While most randomized studies of using this so-called priming strategy have been negative, one important large prospective trial conducted by the HOVON group suggested pre-administration of granulocyte colony-stimulating factor (GCSF) would allow remission to occur at a lower level of tumor burden, thereby leading to superior long-term and overall survival rate.

The standard of care remains 3 days of anthracycline, 7 days of cytarabine, with reinduction in the event of positive leukemic cells in the adequately cellular marrow obtained 15 to 22 days after the start of induction.

For those who fail to achieve remission after one or two cycles of induction chemotherapy, the prognosis is poor. These so-called refractory patients are often salvaged with a high-dose ara-C based regimen. If the salvage attempt is successful, then they should be taken to allogeneic stem cell transplant using the best available donor, even if that is an alternative donor such as haplo-identical matched donor or a cord blood unit or units.

For those who achieve remission, several potential strategies can be undertaken to try to reduce the leukemic burden down to a level compatible with cure. These strategies include a chemotherapy-based approach, a high-dose chemotherapy approach with autologous stem cell rescue, or allogeneic stem cell transplantation. Chemotherapy is thought optimal if it includes high-dose ara-C.

A seminal trial conducted by the CALGB (CALGB 8525) 2 to 3 decades ago, randomized patients who achieved remission after a "3+7" based regimen to four cycles of high-dose ara-C (3g/m2 over 3 hours given every 12 hours on days 1, 3, and 5, intermediate-dose ara-C (400mg/m2 per day for 5 days by continuous infusion), or a lower-dose approach ara-C (100mg/m2 by continuous infusion for 5 days).

These regimens were then followed by a lower-dose regimen involving 1 day of daunorubicin and 5 days of subcutaneously administered low dose ara-C. This trial demonstrated that patients under age 60 who were randomized to the high-dose arm enjoyed a superior disease-free and overall survival, compared to those randomized to the lower-dose arms. Importantly, this disease-free and overall survival benefit was only seen in patients under age 60.

Moreover, subsequent studies revealed that patients who benefitted the most from the intensified high-dose ara-C were those who had an inversion 16 or t(8;21) chromosomal abnormality, the so-called CBF or core binding factor leukemias. While subsequent studies have failed to demonstrate that high-dose ara-C is superior to other intensive regimens and that the very high doses used in CALGB 8525 are required, the overall point is that if chemotherapy is chosen, it should be given intensively. The value of maintenance therapy is debated in AML and is not generally used, although one trial performed in Germany did suggest a benefit.

While high-dose chemotherapy with autologous stem cell rescue has never been demonstrated to be superior to several cycles of an intensive chemotherapy regimen, proponents argue that the total time of myelosuppression is actually lower with a single cycle of the so-called autologous transplant. On the other hand, it may be more difficult to salvage patients with an allogeneic transplant if they relapse after high-dose chemotherapy with autologous stem cell rescue, than if they were treated with an intensive chemotherapy-based approach.

A very important post-remission strategy is allogeneic stem cell transplant. It is clear that relapses after such therapy are much lower than they are after chemotherapy or high-dose chemotherapy with autologous stem cell rescue. Treatment-related mortality is much higher due to graft-versus-host disease, infection, and other complications. Prospective studies, in which patients with AML in remission were allocated to allogeneic stem cell transplant if they had a sibling donor, compared to chemotherapy with autologous stem cell transplant were inconclusive, but failed to show a clear-cut benefit to allogeneic stem cell transplantation in most cases.

However, recent meta-analyses of all such trials suggest that the strategy of performing an allogeneic transplant in first remission for younger adults with AML may be slightly superior to a chemotherapy-based approach, except in patients with chromosomal abnormalities (inversion 16 and t[8;21]) generally portending a good prognosis when chemotherapy is used. If chemotherapy alone is used, the number of cycles to be administered is not clear-cut, although it seems that at least three and possibly four cycles (for patients with good prognosis chromosomal abnormalities) are optimal.

Attempts to maximize graft versus leukemia without worsening graft-versus-host disease (GVHD) have been undertaken. Graft T-cell depletion approaches have been used to minimize GVHD, but the relapse rate is higher. Currently, most centers employ non-T cell depleted grafts from allogeneic donors, but use increasingly more effective GVHD prophylaxis strategies and current antibiotics to minimize post-transplant complications. While allogenic transplantation in first remission AML was formerly only considered viable if a sibling donor existed, much emerging data suggests that a fully molecularly matched unrelated donor yields the same likelihood of overall success as does a fully matched sibling donor.

A so-called risk adapted approach is generally undertaken for patients with AML under age 60 in first remission. Patients with favorable chromosomal abnormalities (inversion 16 and 8;21) are generally treated with four cycles of a high dose ara-C based regimen. An important emerging caveat is that patients in these cytogenetic subgroups who have a c-Kit mutation (approximately 25%) have a poor prognosis and should be considered for an allogeneic stem cell transplant in first remission. The 15% of younger adults with AML who have a so-called adverse prognostic karyotype at diagnosis, should definitely be treated with an allogeneic transplant during first remission.

The donor for this transplant could certainly be a sibling if available, or a completely unmatched individual; however, emerging data from single institutions suggest that cord-blood transplant or even a haplo-mismatched approach might be of sufficiently low risk to justify this approach for an AML patient in first remission with a clearly poor prognosis. The vast majority of AML patients have a normal karyotype or karyotype associated with an intermediate prognosis.

It is becoming the standard of care to separate subgroups of patients with normal karyotypes into those with various gradations of prognosis. Patients with an FLT3 ITD mutation are considered to be at high risk and some, but not all studies have suggested that such patients should be allocated to allogeneic stem cell transplant in first remission. A subgroup that does fairly well with chemotherapy and does not require an allogeneic transplant in first remission are those normal karyotype patients who do not have an FLT3 ITD mutation, but do have an NPM1 mutation. Patients with neither an NPM1 nor an FLT3 mutation are also probably best allocated to allogeneic stem cell transplantation. In the case of a bi-allelic CEBP-a mutation, the prognosis is also quite good, so a chemotherapy-based approach is also indicated.

APL of any age

About 10% of patients with AML have an entity termed, acute promyelocytic leukemia (APL). This subtype is treated differently than the other categories of AML. Most patients with APL present with leukopenia, profound thrombocytopenia, and evidence of DIC. The DIC is believed to be on the basis of release of procoagulant granules, which turn on the clotting cascade and/or activate fibrinolysis producing consumption of fibrinogen; life threatening cerebral hemorrhages are not uncommon during the presentation phase of APL.

The diagnosis is made by noting infiltration of the marrow with blasts and malignant-appearing promyelocytes, which are intensely myeloperoxidase positive and often contain numerous Auer rods. Immunophenotypic studies that would be consistent with the diagnosis of APL include the strong expression of the myeloid antigen CD33 and the absence of the class I antigen HLA-DR. Definitive diagnosis requires the demonstration of the characteristic t(15;17) chromosomal translocation by cytogenetics or FISH, or more specifically by noting the fusion mRNA gene product via PCR.

As soon as the diagnosis of APL is known or even suspected, patients should receive all trans retinoic acid (ATRA 45mg/m2) in two divided doses daily. ATRA will rapidly ablate the DIC and restore hemostasis to a more normal state. While waiting for the DIC to correct, it is appropriate to keep the fibrinogen above 100 to 150 with fresh frozen plasma or cryoprecipitate, and the platelet count greater than 10,000 to 20,000 with the use of platelet transfusions.

While APL can have devastating initial manifestations, this is probably the most curable subtype of AML. A presentation with a WBC greater than 10, 000/ul indicates high risk APL.

A historically popular and effective regimen is the so-called PETHEMA regimen from Spain which uses idarubicin plus ATRA (all trans retinoic acid) induction, along with three courses of anthracycline/ATRA consolidation, followed by maintenance therapy with ATRA plus 6-mercaptopurine and methotrexate. The cure rate for high-risk patients (those with white counts greater than 10,000) is about 60%, but those with intermediate risk, and low risk patients (those with white counts less than 10,000) have an 85% cure rate.

The care of patients with non-high risk APL (at least in patients' ages 18-70) has been revolutionized by the results of a trial led by the GIMEMA group. In this trial, patients with APL with presenting WBC less than 10K were randomized to the PETHEMA regimen or a chemotherapy-free approach of ATRA plus arsenic trioxide (ATO).ATO (0.15mg/kg/per day) was known to have powerful anti-APL activity as it was highly effective in relapsed patients; the ATRA/ATO combination had been successfully piloted at MD Anderson. The GIMEMA trial demonstrated that well over 85% of patients in either arm were apparently cured, but the results were statistically better and on the whole less toxic in the ATRA/ATO arm.

The United States Intergroup (CALGB 9710) demonstrated that the use of arsenic early in the post-remission setting leads to a high rate (80%) disease-free survival. The induction regimen consists of ATRA, plus 4 days of anthracycline, plus seven days of cytarabine induction; post remission therapy included two 25 day cycles of arsenic trioxide, followed by two cycles of daunorubicin/ATRA, and this was followed by 1 year of maintenance therapy with ATRA, 6-mercaptopurine and methotrexate. This regimen was quite effective even in those with higher risk disease. As such either C9710 or the use of a high-dose ara-C should be considered in APL patients who present with WBC >10, 000/ul.

For patients with APL who relapse, arsenic trioxide is the most commonly used regimen, even in patients who previously received arsenic. Gemtuzumab ozogamicin (an antibody targeted immunoconjugate) which binds to CD33 is very effective in this disorder but is no longer routinely available due to marketing withdrawal. Novel retinoids are in development.

If second remission is achieved at a PCR negative state, then high dose chemotherapy with autologous stem cell transplant can be effective. If hematologic but not molecular remission is achieved, then allogeneic stem cell transplant should be considered. An emerging problem for patients with a APL is central nervous system disease, which is more common in those who present with a high white count. Some advocate the use of prophylactic intrathecal therapy in those who present with high-risk APL.

Older adults with non-APL AML

Older adults have a worse outcome than younger adults with AML. The therapeutic outcome in younger adults has improved steadily in recent decades probably due to the use of more intensive chemotherapy,as well as better supportive care. Unfortunately, the same is not true for older adults, who have a remission rate in response to standard chemotherapy of between 40 to 50%; furthermore, the chance for long term survival in those who do achieve remission is less than 20%, yielding a long term disease free survival rate of less than 10%. The median survival of an older adult with AML is approximately 11 to 12 months.

The reasons for the inferior outcome in older adults can be considered either due to host factors or disease-related factors. Older adults have a relatively impaired stem cell reserve, decreased ability to clear chemotherapy due to decline in hepatic and renal function, as well as a tendency to have a significant number of co-morbid diseases. Secondly, the leukemias that tend to arise in older adults are intrinsically resistant, probably arising from a more proximal stem cell in the hematopoietic hierarchy. Several lines of evidence support this notion. Firstly, the ratio of unfavorable to favorable chromosomal abnormalities is higher in older adults than in younger adults. Secondly, leukemic cells from older patients tend to be more likely to express proteins that mediate chemotherapy resistance such as the MDR1 (multidrug resistance protein 1) pump. Thirdly, older patients are more likely to have had a prior stem cell disorder such as myelodysplastic syndrome arising either de novo, or following exposure to leukemogenic chemotherapy. The common finding of mutations in genes which encode for proteins which are epigenetically active, even in patients with no known prior history MDS, suggest that occult stem cell disorders are common in this age group (and similar abnormalities have been noted in aging normal adults).

As in younger adults, the biology of AML in older adults is heterogeneous, with regards to chromosomes and genetics; however, the difference between the outcome in "favorable patients" and those with unfavorable biological features is less pronounced than it is in younger adults. The older adult with an NPM1 mutation who does not have an FLT3 mutation, or an older adult with a core binding factor leukemia, might expect to have as much as a 30% chance for long-term disease free survival, about half the likelihood of good outcome that one would expect in otherwise matched younger adults.

There is a particularly large degree of heterogeneity in terms of host factors based on performance status, co-morbid diseases, and patient age. While performance status may be a very imprecise way of determining suitability for chemotherapy, a more comprehensive geriatric assessment can be performed to yield more accurate information. Numerous risk factor models have been proposed to estimate the likelihood of achieving complete remission in a given older patient with standard chemotherapy. There are many older adults, such as those over age 70, those with poor performance status, those with non-favorable chromosomes, or those with significant co-morbid disease burden who would be expected to have a relatively low likelihood of achieving remission with standard "3+7".

As such, there is much interest in substituting less intensive therapy for the "3+7" backbone, which carries as much as a 20% mortality in this age group. Single agents such as clofarabine, azacitdine and decitabine have undergone testing in untreated older adults with non-APL AML. Remission rates between 25 and 45% have been seen in most cases, with a significant lower treatment related mortality than would be expected with a standard induction chemotherapy. However, whether or not the duration of the remissions and the overall outcome are changed using these less intensive approaches remains to be determined.

The optimal approach to the post-remission setting for older adults with non-APL AML is even less clear. It has been demonstrated that the use of intensive post-remission chemotherapy is not warranted in this age group due to a high degree of morbidity and mortality, as well as a lack of superior outcome compared with the lower dose approach. On the other hand, if an older adult with AML who received induction chemotherapy is to have even a small chance for long-term disease free survival, repeat induction or some less intensive post-remission strategy (compared with that used in younger adults) should be administered.

Presumptively, one way to use the lower dose options in the post-remission setting is simply to repeat (at least twice) the clofarabine or the decitabine that was given as an induction attempt. Conversely, it is now common to refer many older patients who are in remission, particularly those between ages 55 and 75 years for a non-myeloblative (also known as reduced intensity conditioning) allogeneic stem cell transplant, if a suitable donor can be found. Studies have suggested that such reduced intensity allogeneic transplants lead to an approximately 35% likelihood of long term disease free survival in older adults with non-APL AML. Of course, these are highly selected patients.

In one study from MD Anderson, only 10% of all patients in this age group were referred to a stem cell transplant doctor, and even fewer received a transplant. Prospective trials comparing chemotherapy to non-myeloblative allogenic stem cell transplant in this age group are probably not going to happen, but certainly this remains an interesting option for reasonably fit older adults who are willing to take the upfront and post-transplant risks, in an effort to achieve a higher likelihood of long term disease free survival.

When AML relapses after a period of remission, the outcome is poor. A host of "salvage regimens" are available, most notably mitoxantrone, etoposide, cytarabine, or clofarabine/cytarabine. The most important feature of predicting outcome in the relapse setting is the duration of the initial remission, with longer remission portending a better outcome. In general, if a patient is fit enough to undergo salvage chemotherapy, then a good response should be followed by allogeneic stem cell transplant; some patients will turn out to be long-term disease free survivors if they have responsive disease.

What other therapies are helpful for reducing complications?

Nausea and malnutrition

The management of chemotherapy-induced complications is one of the most important factors promoting the successful treatment of patients with AML, particularly in younger adults receiving induction and/or consolidation chemotherapy. First, modern antibiotic therapy has revolutionized the care of patients with AML. The use of serotonin agonists such as ondanesetron has markedly reduced the nausea and vomiting associated with the chemotherapy used in AML. It is important to make sure that patients are well hydrated, as their oral intake inevitably drops.

Tube feedings are not generally used because of the concern that indwelling tubes in patients with neutropenia and thrombocytopenia will promote infection and bleeding. The period of malnutrition is not usually sufficiently prolonged to require IV nutrition. Total parenteral nutrition given by a central line can supply calories but is costly, is associated with an increased risk of fungal and bacterial infections, and is not of proven benefit in affecting outcome.

Managing skin and gut flora infections

Patients with AML who sustain profound and prolonged neutropenia are at risk for infections with commensal organisms such as skin and gut flora. Benefit from specialized air handling is debatable. Our practice is to insure that visitors wash their hands and to have patients in a positive pressure environment. Should patients who are neutropenic after AML chemotherapy eat fresh fruits and vegetables? Some data suggests that food intake can be liberalized; but so-called nutropenic diets that remove fresh fruits and vegetables are commonly used.

Managing tumor lysis syndrome

A potential complication of AML is tumor lysis syndrome, spontaneous or chemotherapy-induced cell lysis, leading to the release of large amounts of potassium phosphate causing hyperkalemia, hyperphosphatemia, and secondary hypocalcemia. Purine metabolites lead to uric acid accumulation in renal tubules which can cause renal failure. Although most AML patients do not have profound tumor lysis, provision of IV normal saline and allopurinol (an xanthine oxidase inhibitor) to prevent renal deposition of uric acid is worthwhile. Incipient renal failure and/or an inability to take oral medicines should prompt the use of recombinant urate oxidase to rapidly decrease the serum uric acid.


Some AML patients, particularly those with acute monocytic or acute myelomonocytic AML who present with very high white counts, can develop the phenomenon of leukostasis. This is caused by adherent myeloblasts leading to microhemorrhages and/or microthrombi in the cerebral or pulmonary vasculature. When a patient with AML has a blast count exceeding 20,000 to 30,000, providers should at least consider the possibility that leukostasis might occur. Most patients can be managed with IV fluids and hydroxyurea, an orally administered ribonucleotide reductase inhibitor, which can easily control the white count.

More complicated maneuvers such as leukopheresis (temporizing measures that must be followed up with definitive anti-leukemic therapy) are rarely needed. Leukopheresis is generally reserved for patients in whom hydroxyurea cannot be used, or is not working well and who have incipient leukostatic complications. A rarely used adjunctive strategy is so-called “throwaway” radiation in which a brief course of radiation to the brain or lungs rapidly lyses tumor cells in emergency situations.

Managing infections

By far, the most important complication in AML patients is infection. A full discussion of the anti-microbrial management of the AML patient is beyond the scope of this treatise. Broad spectrum antibiotics such as ceftazidime are the mainstay of the empiric treatment of fever and neutropenia in such patients. Antistaphylococcal antibiotics such as vancomycin are usually reserved for patients who have demonstrated or presumed skin infections. Empiric antifungal therapy with micafungin or a similar agent should be initiated in the event of persistent fever in the face of broad spectrum antibiotics in neutropenic patients. The antifungal azoles are used by many centers earlier in the course as prophylaxic agents or in the case of a micafungin resistant infection. The use of effective and relatively non-toxic broad spectrum antimicrobial agents has been a key factor in the improved outcome of younger patients with AML in the past 20 years.

Psychosocial support

Finally, it is extremely important to tend to the psychosocial needs of the patient and his/her family. The diagnosis of acute leukemia is a devastating event in the patient’s life. At the very least, it will mean months of difficult chemotherapy, and a possible allogeneic stem cell transplant. The latter means obligate infertility. Consequently, sperm banking or ovarian preservation should be considered at the time of diagnosis and at first remission for younger patients.

Moreover, since the vast majority of patients with AML die of their disease, end-of-life issues can become of paramount importance. Every patient with AML should be seen at diagnosis by a social worker and/or psychiatrist. In this context, it is important to emphasize the team approach to the patient with AML as input from medical oncology, psychiatry, social work, pharmacy, nutrition, and physical therapy are all needed to optimally care for such complicated patients.

What should you tell the patient and the family about prognosis?

The prognosis of patients with AML depends largely on age and cytogenetics. Age is a critical prognostic factor, since patients older than age 55 to 60 are rarely cured of this disease, yet patients under age 40 can expect a 40% cure rate. The reason for the inferior outcome in older adults, as explained above, has to do with both host and disease factors.

AML in older adults tends to be a more intrinsically resistant disease, often presenting after a prior hematological abnormality and/or after prior exposure to potentially leukemogenic chemotherapeutic agents for other cancers. Secondly, older adults have more difficulty tolerating the intensive chemotherapy used to treat AML, due to decreased stem cell reserve, as well as decreased ability to clear chemotherapeutic agents, given relatively impaired liver and kidney function.

Especially in younger adults, the chromosomal analysis in the leukemic blasts offers critical prognostic information. About 15% of patients have an adverse chromosomal pattern at diagnosis (often involving loss of the long arm or all of chromosome 5 and/or 7, and/or complex cytogenetic abnormalities). Such patients and especially those with the ‘monosomal’ subtype (to monosomies or one monosomy and one structural abnormality) have a dismal prognosis (less than 20% cure rate with chemotherapy) and should be referred for allogeneic transplantation. Another 15% of patients have what is considered a favorable prognosis if chromosome studies reveal either an inversion of chromosome 16, the translocation t(8;21), or the t(15;17) translocation typical of acute promyelocytic leukemia. The inversion 16 and t(8;21) chromosomal abnormalities are considered core-binding factor leukemias, because in each case one partner of the balanced translocation encodes the protein involved in the core binding factor transcription heterodimer. Such patients can expect to have about a 50% cure rate and as high as 85-95% for the APL patients.

However, about 25% of patients with core-binding factor leukemias have an activating mutation in the c-Kit tyrosine kinase gene which portends a poor prognosis. About 70% of patients with AML have intermediate prognostic chromosomal abnormalities. A very large body of recent work has begun to parse patients with normal karyotype AML into differing prognostic categories.

For example, patients with an NPM1 mutation who do not have an FLT3 ITD mutation, have a fairly good prognosis, and might not require allogeneic stem cell transplant as a consolidation therapy; whereas a patient who has an FLT3 ITD mutation, or those without either an FLT3 ITD or an NPM1 mutation have an inferior prognosis and might be better served by allogeneic stem cell transplantation in first remission. Finding mutations in other genes such as IDH1, IDH2, TET2,and DNMT3A may be able to find refine prognosis in normal karyotype individuals. A fairly rare mutation which carries a good prognosis is a bi-allelic mutation in the CEBP-a transcription factor.

For most older adults, the outcome is quite poor. Approximately 40 to 50% of fit older adults receive induction chemotherapy and they have a median overall survival of about 11 months. Only about 40% of these achieve remission, of whom less than 20% remain long-term disease free survivors. Within the older group, there are certain prognostic factors that predict for an even poorer outcome with induction chemotherapy. Such factors include age greater than 70, ECOG (Eastern Cooperative Oncology Group) performance status less than 1, co-morbidities, history of antecedent hematologic abnormality or non-favorable chromosomes.

In fact, only a relatively small number (under 10%) of older adults present with favorable chromosomes (those involving the core binding factor transcription factors apparatus or APL). Aggressive chemotherapy may be appropriate in selected older adults. For example, those with a normal karyotype, but who have an NPM1 mutation without a concomitant FLT3 mutation have a 30% long-term disease free survival.

In APL, the prognostic factors are related simply to the platelet count and the white count at the time of diagnosis. The best combination at presentation is a white count less than 10,000 and a platelet count greater than 40,000, and the worst are those with high risk disease (the 10 to 15% of APL patients who present with a platelet count of less of than 40,000 and a white count of greater than 10,000). Such patients don’t do as well when given an anthracycline/retinoic acid based regimen, but may do much better when arsenic trioxide is applied early in the treatment course.

What if scenarios.

Use of intensive post-remission chemotherapy in the treatment of older adults with non-APL AML

Most clinicians realize that intensive doses of ara-C or similarly intensive combinations of other standard anti-leukemic regimens are better than lower doses of chemotherapy in the post-remission management of adults with AML. However, these results cannot be extrapolated to the treatment of older adults who achieve remission with standard induction chemotherapy.

Unfortunately, dose intensive therapy has never been shown to be useful in the older patient cohort. An exception to this rule could be those rare patients who present with a core binding factor leukemia or with an NPM1 mutation without an FLT3 mutation in the setting of a normal karyotype AML. Even these groups don’t do all that well with chemotherapy, but probably would benefit from intensive chemotherapy if it can be feasibly applied. For the vast majority of older adults, intensive post remission chemotherapy is neither effective, or well tolerated.

On the other hand, the optimal post remission strategy in this cohort is not known. A reasonable approach is either to repeat the induction or use modified high dose ara-C (that is, 1gm/m2 daily for 5 days for one or two cycles). Ironically, although intensive chemotherapy is not recommended, it is reasonable, based on current data from single institutions, to allocate older patients with AML in first remission to reduced intensity conditioning allogenic stem cell transplant. Reports (non-randomized) have suggested that such patients might enjoy a 30 to 40% disease free survival.

Waiting for clearance of infections before beginning definitive antileukemic therapy in a newly presenting patient

Infections such as pneumonia or cellulitis can be the presenting manifestation in an AML patient; appropriate antibiotics should be promptly administered. However, it makes little sense to wait until the infection improves before beginning anti-leukemic chemotherapy. Even though such therapy will cause a profound drop of the neutrophil count, without clearance of the blasts and resumption of normal hematopoiesis, the infection will not be cleared; consequently, as soon as the diagnosis is made, effective anti-leukemic chemotherapy should be administered while antibiotics are being given, even when the infection is not completely clinically resolved.

Improper use of day 15 bone marrow when using standard induction chemotherapy

The standard way to apply "3+7" induction in adults with AML is to perform a mid-cycle or day 15 marrow, to determine if the original chemotherapy was successful enough to have ablated the marrow and reduced the number of leukemic cells. The day 15 marrow can be reasonably done between day 15 and 22, although in most cases performing this procedure in the earlier part of this time frame is more appropriate.

To diagnose persistent leukemia, one should see at least 20% blasts in a marrow that is at least 10% cellular. If any doubts exist about the enumeration of the cellularity or blasts to meet this criteria, a repeat marrow should be done between 5 and 7 days later, to clarify the situation. One should also review the peripheral blood smear, especially when the initial mid-cycle or repeat mid-cycle marrow is being done about 3 weeks after the start of the "3+7" induction.

At that point, one might see early recovery with increasing numbers of monocytes, mid-range myeloid cells, and a higher platelet count, even if some blasts are present in the differential. In this setting, bone marrow blasts might be increased, but these might be normal blasts. As such, reinduction therapy with a repeat course of "3+7" or with so-called "2+5" therapy should wait until it can be clarified whether the blasts represent persistent leukemia cells or normal recovering hematopoietic cells. This dilemma can be solved either by waiting to observe the pattern of recovery over several days or by using special stains or immunophenotype to identify whether these blasts are part of the same malignant clone with which the patient presented.

Insufficient use of genetic testing at diagnosis

For patients with a core-binding factor leukemia, such as inversion 16 or t(8;21), it is a good idea to test, if possible, for a KIT mutation. Those with a KIT mutation have an inferior prognosis compared to KIT wild-type CBF patients who have a good prognosis. For patients with normal chromosomes, the prognosis can be distinguished by checking for the presence of an FLT3 ITD, NPM1 mutation, and for a bi-allelic CEBP-a mutation.

While initial therapy will not differ based on the results of these, patients with an FLT3 ITD mutation fare poorly and should be considered for an allogeneic stem cell transplant. On the other hand, those with an NPM1 mutation without a concomitant FLT3 mutation, and those with bi-allelic CEBP-a mutation can do well with a chemotherapy-only strategy. Certainly the panel of important genes to be tested will expand with additional research. Already preliminary studies have suggested that detecting a DNMT3A mutation, a TET2 mutation, or an IDH1/IDH2 mutation will also allow modification of the prognosis.

Overuse of leukopheresis

Leukocytosis in AML patients can be a medical emergency, especially in acute monocytic leukemia in which the blasts are adherent and tend to more likely cause microhemorrhage or microthrombi in the cerebral and pulmonary vasculature. Most patients with a high white count AML can be managed successfully with IV fluids and hydroxurea to lower the blast count. Very occasionally, a patient who is unable to take oral medicines and/or has ongoing hypoxemia or confusion thought due to such adherent myeloblasts, might be candidate for leukopheresis; however, leukopheresis is only a temporizing measure, which is not all that effective in reducing the blast burden.


The molecular and cellular biology of AML represents a conglomeration of genetic abnormalities that can lead to failure to undergo apoptosis (programmed cell death), exuberant proliferation, failure to differentiate, autocrine or paracrine elaboration of survival factors, and many other types of abnormalities. Gilliland proposed the two-hit model of leukemogenesis for AML, in which mutations in a gene that promoted proliferation such as a gain-of-function tyrosine kinase mutation (for example, FLT3 ITD), plus a mutation in a gene that would impair differentiation such as a transcription factor (for example RUNX1) would be sufficient for leukemogenesis.

However, with the genome sequencing studies performed by Ley and colleagues at Washington University in St. Louis, we now know that even a de novo patient will have as many as ten mutations at the time of diagnosis. Patients with secondary AML or relapsed AML likely have even more. Some of these mutations have turned out to be recurrent (seen in more than just a handful of AML patients). For example, TP53 mutations, typically are found as ‘sole’ lesions; diminished genome protection/apoptosis promoting TP53 function is associated with complex chromosomal abnormalities and a poor therapeutic outcome. Moreover, the AML genome atlas has divided the genetic complexity into nine subgroups: transcription-factor fusions, cohesin-complex, NPM1, spliceosome, tumor suppressor, chromatin-modifying, transcription factor, DNA-methylating, and signaling.

The pathophysiology of therapy-related AML was thought to involve an increased rate of pro-leukemic mutations engendered by therapy-induced defects in DNA-repair. Recent data suggests: 1) AML cells from some patients who have had exposure to prior cancer therapy have mutations which suggest de novo derivation and 2) TP53 mutations may be present in very minor clones that predate chemo- or radiotherapy exposure and are selected for as a consequence of such therapy.

Especially notable are genes involved in epigenetics, or the ability to change gene expression beyond what is encoded by functional genes. Mutations in IDH1 and IDH2 have been discovered to occur in about 20% of patients with AML. These genes encode proteins that encode the isocitrate dehydrogenase enzyme. Point mutations in these genes encode a neomorphic enzyme which produces an alternative reaction product (2-hydroxyglutarate) that may affect the genome methylation status. Secondly, a mutation in the DNA methyltransferase 3 gene also changes the epigenetic pattern, and thus gene expression. Therefore, it seems likely that the pathophysiology of AML is quite complex.

What other clinical manifestations may help me to diagnose acute myeloid leukemia?

The diagnosis of AML is not complicated and is based on symptoms and signs of bone marrow failure, most notably infection and bleeding. However, it is important to determine if any pre-existing conditions that predispose the patient to AML are present. These would include exposure to chemotherapy for prior cancers, exposure to industrial solvents such as benzene, or exposure to radiation therapy from military, industrial, or therapeutic sources. Furthermore, because many patients with AML will require stem cell transplant, it is important to take a good family history to determine if there are siblings who could potentially serve as donors, should the need arise.

On physical examination, lymphadenopathy and splenomegaly are not that unusual, particularly in the monocytic subtypes of AML. Also in monocytic subtypes of AML, gingival hypertrophy, leukemic infiltration of the skin (leukemia cutis), and meningeal signs from leukemic meningitis (cranial neuropathy due to basilar infiltration) should be sought. Petechiae or microhemorrhages in the skin should be noted.

What other additional laboratory studies may be ordered?

Basic studies:

  • Complete blood count with differential

  • Chemistry panel, including uric acid

  • Bone marrow aspirate and biopsy with cytogenetics and genetics (FLT3 ITD, NPM1, CEBP-a, c-Kit)

  • Chest x-ray

  • PT (prothrombin time), PTT (partial thromboplastic time), fibrinogen

  • HLA type

What’s the evidence?

Appelbaum, FR, Gundacker, H, Head, DR. "Age and acute myeloid leukemia". Blood. vol. 107. 2006. pp. 3481-5.

(Based on review of successive trials from the SWOG (The Southwest Oncology Group), the magnitude of the problem of AML in older adults is defined. Risk factors for particularly poor outcome are noted.)

Grimwade, D, Walker, H, Oliver, F. "The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. The Medical Research Council Adult and Children’s Leukaemia Working Parties". Blood. 1998. pp. 2322-33.

(One of the largest studies showing the key impact of diagnostic cytogenetics on outcome in AML.)

Marcucci, G, Haferlaek, T, Dohner, H. "Molecular genetics of acute myeloid leukemia: prognostic and therapeutic implications". J Clin Oncol.. vol. 29. 2011. pp. 475-86.

(A current view of the therapeutic and prognostic importance of genetic diagnosis in AML, particularly in patients with a normal karyotype.)

Schlenk, RF, Dohner, K, Krauter, J. "Mutations and treatment outcome in cytogenetically normal acute myeloid leukemia". N Engl J Med.. vol. 98. 2008. pp. 1752-9.

(The basis for the current recommendation to perform alloSCT (allogenic hematopoietic stem cell transplantation in CR1 in patients with normal cytogenetics unless an NPM1 mutation is a found in a patient with FLT3 ITD wild type status.)

Rollig, C, Bornhauser, M, Thiede, C. "Long-term prognosis of acute myeloid leukemia according to the new genetic risk classification of the European Leukemia Net recommendations: evaluation of the proposed reporting system". J Clin Oncol.. vol. 29. 2011. pp. 2758-65.

(Details whether the 'new' combined prognostic system which involves both diagnostic cytogenetics and analysis of mutations in selected genes makes sense.)

Ley, TJ, Ding, L, Walter, JH. "DNMT3A mutations in acute myeloid leukemia". N Engl J Med. vol. 363. 2010. pp. 2424-33.

(One of several key studies from T Ley and colleagues who have performed deep total genomic sequencing on AML cells (and paired normal tissue). Among many interesting findings, this paper shows that DNMT3A are recurrent in AMl and auger for a relatively poor prognosis.)

Fernandez, HF, Sun, Z, Yao, X. "Anthracycline dose intensification in acute myeloid leukemia". N Engl J Med. vol. 361. 2009. pp. 1249-59.

(Ninety mg/m2 is the optimal dose of daunorubicin to be given during the first three days of induction therapy for most younger adults with AML.)

Bloomfield, CD, Lawrence, D, Byrd JC, Caroll. "Frequency of prolonged remission duration after high-dose cytarabine intensification in acute myeloid leukemia varies by cytogenetic subtype". Can Res. vol. 58. 1998. pp. 4173.

(High dose ara-C consolidation mainly benefits those with CBF cytogenetics. One of the first papers to document the interplay of biology and therapy in AML.)

Koreth, J, Schlenk, R, Kopecky, KJ. "Allogeneic stem cell transplantation for acute myeloid leukemia in first complete remission: systemic review and meta-analysis of prospective clinical trials". JAMA. vol. 22. 2009. pp. 2349-61.

(Based on a meta-analysis, allogeneic stem cell transplantation is the optimal post-complete remission therapy for most younger adults with AML in first complete remission.)

Powell, BL, Moser, BK, Stock, W. "Arsenic trioxide improves event-free and overall survival for adults with acute promyelocytic leukemia: North American Leukemia Intergroup Study C9710". Blood. vol. 116. 2010. pp. 3751-57.

(Use of (50 days) arsenic trioxide improves survival if employed early in the post-complete remission setting for patients with APL.)

Gert, Ossenkoppele, Bob, Lowenberg. "How I treat the older patient with acute myeloid leukemia". Blood. vol. 125. 2015. pp. 767-774.

(A thoughtful approach guiding the management of older adults with AML, suggesting that aggressive treatment should be considered for virtually all such patients.)

"The Cancer Genome Atlas Research Network. Genomic and Epigenomic Landscapes of Adult De Novo Acute Myeloid Leukemia". N Engl J Med. vol. 368. 2013. pp. 2059-2074.

(This work represents the most complete compendium of somatic mutations in patients with de novo AML; the genetic complexity is subdivided into nine pathophysiologically distinct types.)

Wolach, O, Stone, RM. "How I treat mixed-phenotype acute leukemia". Blood. vol. 125. 2015. pp. 2477-85.

(The treatise is a helpful review of diagnosing mixed phenotype acute leukemia, an entity distinct from AML, one generally best managed with ALL-type chemotherapy followed by allogeneic stem cell transplant.)

Lindsley, RC, Mar, BG, Mazzola, E. "Acute myeloid leukemia ontogeny is defined by distinct somatic mutations". Blood. vol. 125. 2015. pp. 1367-76.

(A detailed study of the genetic lesions underlying secondary AML and therapy-related AML; there are clear implications for the therapy of older patients with apparently de novo AML, a subset of which have de novo ( also called ‘pan-AML’) mutations and do fairly well with chemotherapy but those with either secondary-type or TP53 mutations fare poorly with standard chemotherapy.)

Jaiswal, S, Fontanillas, P, Flannick, J. "Age-related clonal hematopoiesis associated with adverse outcomes". N Engl J Med. vol. 371. 2014. pp. 2488-2498.

(Generally small clones of cells with mutations associated with MDS and AML exist is the blood cells of a sizeable minority of clinically normal older adults; patients with such clones have an inferior life expectancy that those without them, but the reason may be cardiovascular events, not hematologic neoplasms.)

Klepin, HD, Geiger, AM, Tooze, JA. "Geriatric assessment predicts survival for older adults receiving induction chemotherapy for acute myelogenous leukemia". Blood. vol. 121. 2013. pp. 4287-94.

Klepin, HD, Rao, AV, Pardee, TS. "Acute myeloid leukemia and myelodysplastic sundromes in older adults". J Clin Oncol. vol. 32. 2014. pp. 2541-2552.

(Two papers detailing useful considerations regarding the approach to the management of older adults with AML including employing the comprehensive Geriatric Assessment battery that assesses multiple dimensions including functional capacity, cognition, and social factors to gage the likelihood of a successful outcome with chemotherapy.)

Lo-Coco, F, Avvisati, G, Vignetti, M. "Retinoic acid and arsenic trioxide for acute promyelocytic leukemia". N Engl J Med. vol. 369. 2013. pp. 111-21.

(A chemotherapy-free approach using all-trans retinoic acid and arsenic trioxide leads to a 92% cure rate (superior to anthracycline/retinoic acid-based therapy) has established a new standard for the treatment of acute promyelocytic leukemia patients who present with WBC < 10K/ul.)

Grimwade, D, Freeman, SD. "Defining minimal residual disease in acute myeloid leukemia: which platforms are ready for “prime time”?". Blood. vol. 124. 2014. pp. 3345-3355.

Kayser, S, Schlenk, RF, Grimwade, D. "Minimal residual disease-directed therapy in acute myeloid leukemia". Blood. vol. 125. 2015. pp. 2331-5.

(The assessment of minimal residual disease in AML to determine the depth of remission my offer useful prognostic information and could be used to guide subsequent therapy.)

Rollig, C, Ehninger, G. "How I treat hyperleukocytois in acute myeloid leukemia". Blood. vol. 125. 2015. pp. 3246-3252.

(We are offered a practical approach to the managment of high WBC, and the possible complications thereof, in AML.)
You must be a registered member of ONA to post a comment.

Sign Up for Free e-newsletters

Regimen and Drug Listings


Bone Cancer Regimens Drugs
Brain Cancer Regimens Drugs
Breast Cancer Regimens Drugs
Endocrine Cancer Regimens Drugs
Gastrointestinal Cancer Regimens Drugs
Genitourinary Cancer Regimens Drugs
Gynecologic Cancer Regimens Drugs
Head and Neck Cancer Regimens Drugs
Hematologic Cancer Regimens Drugs
Lung Cancer Regimens Drugs
Other Cancers Regimens
Rare Cancers Regimens
Skin Cancer Regimens Drugs