The Art of Mixing Follicular Units and Follicular Groupings in Hair Restoration Surgery
 
DOMINIC A. BRANDY, MD
 
Clinical Instructor, Department of Dermatology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania

Dermatologic Surgery. Volume 30. Number 6. June 2004. ISSN 1076-0512
 
BACKGROUND. Follicular grafting yields good results, but takes an excessively long rime to perform the procedure. Iatrogenic damage can also occur to the follicular units during the dissection phase when two follicular units are very close together. OBJECTIVE. The objective was to minimize the length of time to perform the procedure and to limit iatrogenic damage to the follicular units during the dissection phase.
METHODS. A donor strip is divided under a stereoscopic microscope into one-, two-, three-, and four-haired follicular units and three- and four-haired follicular groupings. Three- and four-haired follicular groupings are fanned when 2 two-haired follicular units or a three-haired follicular unit and one-haired follicular unit an, very close together. These hair grafts are then inserted into incisions that are 1.0 to 2.5 mm in length and are arranged in accordance to the particular hairstyle desired by the patient.
RESULTS. The technique used in this article accomplisher, consistently excellent results while creating an environment that prevents damage to follicular units that are extremely close together. This technique requires less time and damages fewer follicular units.
CONCLUSIONS. When performing hair restoration surgery, it " many times counterproductive to use follicular units exclusively.
 
DOMINIC A. BRANDY. MD HAS INDICATED NO SIGNIFICANT INTEREST WITH COMMERCIAL SUPPORTERS.
 
HAIR RESTORATION surgery has gone through a vast array of vacillating changes over the past 50 years. Originally, in 1939, a Japanese dermatologist named Okuda^I introduced the use of small full-thickness autografts of hair-bearing skin for the restoration of the scalp eyebrow, and mustache areas. In 1943, Tamura² published an article on the use of single hairs to reconstruct the female pubic area. Subsequently, Fujita³ used a similar technique to treat scars, eyebrows, eyelash defects, pubic areas, and alopecia areata. But because these original articles were published in the Japanese literature and World War II was under way, they went unnoticed by the rest of the world. In the early 1950s, Orentreich⁴ with no previous knowledge of the Japanese experience, began experimenting with the concept of donor and recipient dominance. He took biopsies of normal skin and diseased skin and transplanted these into other areas of the body to evaluate whether or not the biopsied skin would maintain its characteristics or would change in accordance to the area it was grafted into. He also took biopsies of scalp on the sides of the head of balding men and transplanted these into balding areas. What he found was that the scalp from the occipital and parietal areas would grow and thrive in an area where hair was no longer growing. Thus, the theory of donor dominance was conceived-the donor hail would dominant in an area of pattern baldness. It was this revelation that catapulted the field of hail restoration surgery in the United States.
 
During his initial experience, Orentreich would create the grafts with 6.0- to 12-mm trochars and would perform six-punch grafts (Berger R, personal communication, July 1995) at a time. But as the years went on it was not unusual for he and other practitioners^5-7 to perform one-hundred 4.0-mm punch hair grafts per session. This technique resulted in an average of 20 to 30 hairs per graft, but the results were usually cosmetically unpleasing because of the fact that the 4.0-mm plug had no counterpart in nature. Hair restoration surgery made a major leap in 1981 when Nordstrom and colleagues⁸ introduced the idea of utilizing three- to six-haired grafts at the anterior edge of the hairline. Later, in 1984, two major events occurred that would change the face of the field forever-Bradshaw⁹ introduced the idea of utilizing smaller six- to eight-haired quarter grafts over the entire head to achieve more refined cosmeses and Headington¹⁰ clarified the follicular unit. In Headington's article the medical community learned that a follicular unit included:

• One to four terminal follicles;
• One (or rarely two) vellus follicles;
• Nine associated sebaceous lobules;
• Insertions of the arrector pili muscles;
• Perifollicular vascular plexus;
• Perifollicular neural net; and
• Perifolliculum-circumferential band of fine adventitial collagen that defines the unit.

This definition helped us understand that the follicular unit was not only an anatomic unit, but was more importantly a physiologic one. This understanding subsequently ushered in the philosophy that the follicular unit should be left intact and not be broken apart. In 1987, Limmer¹¹ was the first to use Headington's information clinically by utilizing the stereoscopic microscope to create follicular units, which would later be used for surgical hair restoration. Later in 1996, Seager¹² gave significant credence to the philosophy of keeping the follicular unit intact when he demonstrated that when single-hair micrografts were created from follicular units, their growth was not as good as intact follicular units. In his study, he demonstrated that at 5 1/2 months the one-haired micrografts had an 82% survival rate, compared to the 113% survival rate of intact follicular units. This 113% rate was most likely due to the sudden growth of anagen effluvium hairs added to the new growth from follicular units.
 
Since 1996, the exclusive use of follicular units has gained in popularity primarily owing to the writings and lectures of Rassman and Bernstein^13,14 and others. Although results with the exclusive use of follicular units is physiologically sound and aesthetically pleasing, it does have the downfall of taking an inordinately long time (6-8 hr) to complete a session and being more costly to the patient. There also can be some follicular damage when two follicular units are extremely close together. These two negatives are especially evident when the cosmetic surgeon performs follicular unit transplantation infrequently, because the staff does not have the adeptness to move quickly. Although some centers report transection rates of 0% to 3%, these are among the very elite facilities and do not represent the average hair restoration surgeon. This article demonstrates a technique that leaves the follicular unit intact, but allows for the use of three and four-haired follicular groupings when follicular units are very close together. This combination allows for a quicker procedure and less iatrogenic damage of follicles, both of which should give better hair growth, and is less costly to the patient.
 
Graft Creation Technique
A 1.0 x 20.0-cm fusiform strip is normally excised from the donor site and is immediately handed to the surgical assistants who begin slivering the strip into very fine strips that are one follicular unit in width with the help of a stereoscopic microscope (Figure 1). A multiblade strip technique is not utilized because of the known increase in iatrogenic damage to the follicles during extraction. Once the very thin slivers from the fusiform strip are completed, the assistants then dissect these into six different types of grafts:

• One-haired follicular units;
• Two-haired follicular units;
• Three-haired follicular units;
• Four-haired follicular units;
• Three-haired follicular groupings; and
• Four-haired follicular groupings.

A one-haired follicular unit is obviously a single hair egressing the scalp alone (Figure 2). A two-haired follicular unit is a bundle of two hairs egressing the scalp through one opening (Figure 3). A three-haired follicular unit is similarly a bundle of three hairs egressing the scalp through one opening (Figure 4). A four-haired follicular unit is a bundle of four hairs egressing the scalp through one opening (Figure 5).

 
Figure 1. A 1.0 x 20-cm fusiform strip is normally excised from the donor site and is immediately handed to the surgical assistants who begin slivering the strip into very fine strips that are one follicular unit in width.

Figure 2. A one-haired follicular unit is a single hair egressing the scalp alone.	Figure 4. A three-haired follicular unit is a bundle of three hairs egressing the scalp through a single opening.
Figure 3. A two-haired follicular unit is a bundle of two hairs egressing the scalp through a single opening.	Figure 5. A four-haired follicular unit is a bundle of four hairs egressing the scalp through a single opening.

Although Bernstein¹⁵ has written that the average patient as one follicular unit per square millimeter, these follicular units are not as equally distributed as one might think. There are many times when two follicular units are much closer together than surrounding follicular units (Figure 6). In these cases, the author diverges from the follicular purist and makes a follicular grouping instead of two follicular units. A three-haired follicular grouping is made when a two haired follicular unit and a one-haired follicular unit are so close together that the assistant feels that there will be a high probability of iatrogenic injury if the two follicular units were dissected apart. In this case, the assistant incorporates the two follicular units, which will make a follicular grouping that will obviously have more tissue (Figure 7). This follicular grouping will therefore require a larger incision at the time of preparation of the recipient sites. A four-haired follicular grouping is formed when either a two-haired follicular unit and another two-haired follicular unit are extremely close together (Figure 8) or a three haired follicular unit and a one-haired follicular unit are juxtaposed (Figure 9).

Figure 6. Photograph demonstrating that there are times when two follicular units are much closer together than the surrounding follicular units.	Figure 8. A four-haired follicular grouping can be created when two-haired follicular units are extremely close together.
Figure 7. A three-haired follicular unit is created when a two-haired follicular unit and a one-haired follicular unit are so close together that there is a high probability of iatrogenic injury if the two follicular units are dissected apart.	Figure 9. A four-haired follicular grouping can also be created when a three-haired follicular unit and one-haired follicular unit are extremely close together.

Another type of hair graft that has not been written about much in the medical literature but has been observed frequently at the author's clinic is a three or four-haired bundle that appears to be a follicular unit at the surface, but has hair shafts that bow away from the center as the hair enters deeper toward the matrix (Figure 10). Because these grafts require preservation of a fair amount of tissue, which requires a larger incision, we classify these as follicular groupings. 

Figure 10. Occasionally, one will observe three or four hairs exiting the scalp in a bundle, but as the hair shaft enters the deeper scalp, the roots bow away from the center. Because those grafts require a large incision, the author classifies those as follicular groupings.

Incisional Technique
The author uses three different spear blades for all six different grafts. A 2.0-mm Swann Morton No. 91 is usually used for all three- and four-haired follicular groupings. A 1.5-mm Swann Morton No. 90 is typically utilized for all two-, three-, and four-haired follicular units and an Ellis 1.0-mm spear blade is normally used for all one-haired follicular units. Although these blades are usually used in the aforementioned fashion, the author will perform test incisions for each of the grafts so that a perfect snug fit ensues.
 
The advantage of using only a 1.5-mm Swann Morton No. 90 for two-, three-, and four-haired follicular units (instead of a different blade for each) is that the final decision as to where the follicular units will be placed can be slightly altered after all the incisions are completed. This is important because sometimes the prediction as to the breakdown of the various follicular units will change as the assistants are cutting the grafts. Typically, the author knows the exact breakdown of the various grafts when the staff is one-half way through dissecting the fusiform strip, but this does not always occur.
 
When using 2.0-mm Swann Morton No. 91 blades for three- and four-haired follicular groupings, the author averages 8 incisions/cm², which yields approximately 30 hairs/cm². Conversely, when Swann Morton No. 90 incisions are made for two-, three-, or four-haired follicular units, the author normally makes approximately 20 incisions/cm². If two-haired follicular units were being placed, these incisions would yield approximately 40 hairs/cm². If three-haired follicular units are being utilized this would calculate out to be 60 hairs/cm² and if four-haired follicular units are being used there would be a yield of approximately 80 hairs per cm². When using a 1.0-mm Ellis blade for one-haired follicular units, the author usually makes approximately 25 incisions/cm², which yields approximately 25 hairs/cm². It is the author's opinion that it is essential that when working in areas where preexisting hair is growing, that 3.5 x expanded-loupe magnification be utilized while making the incisions so that preexisting hair follicles are not damaged.¹⁶
 
Case Studies
To demonstrate how the cutting and incisional techniques are incorporated on individual patients, the author presents a series of three case studies. The total number of hairs presented in each donor strip is the net number of hairs derived following follicular unit dissection

Case Study 1
The patient is a 50-year-old Oriental man who presents with Type VI male pattern baldness and fine, straight and black hair (Figure 11A). He desires to part his hair from the left and backward. Upon harvesting a 1.0 X 20.0-cm strip the following characteristics were present in the strip:

• 2485 hairs (124 hairs/cm²);
• 1401 follicular units (70 follicular units/cm²);
• 559 one-haired follicular units (40% of all follicular units);
• 600 two-haired follicular units (43% of all follicular units); and
• 242 three-haired follicular units (17% of all follicular units).

The strip was divided into the following grafts:

• 214 one-haired follicular units;
• 255 two-haired follicular units;
• 242 three-haired follicular units; and
• 345 three-haired minigrafts made from 345 two haired follicular units combined with 345 one-haired follicular units.

Figure 11C demonstrates the placement of the aforementioned grafts in the density distributions previously discussed in this article. The following is the breakdown of how the grafts were distributed:

• 90 one-haired follicular units are used for anterior refinement of the hairline.
• 80 one-haired follicular units are scattered peripherally into the superior parietal scalp down to the point where future hair loss is predicted by hair wetting.
• 44 one-haired follicular units are scattered at the posterior hairline abutting the vertex.
• 100 two-haired follicular units are used directly behind the immediate hairline for gradation.
• 155 two-haired follicular units are used centrally to gradate from the part-side heavily packed three haired follicular region to the less dense area of follicular groupings.
• 242 three-haired follicular units are packed densely on the part side to create the illusion of more hair than if the hair was equally placed over the entire area.
• 345 three-haired follicular groupings are placed into the nonpart side where the density does not need to be as intense.
  

After two sessions utilizing this approach the results are as shown in Figure 11B. One can notice that although only 4920 hairs were utilized in this patient, a good result ensued primarily owing to the fact that the hairs were strategically placed in accordance with the desired hairline. Obviously, if this patient wanted further enhancement of the density another session could be performed. 

Figure 11. (A) The patient is a 50-year-old man with Type VI baldness from the frontal view, preoperatively. (B) Same patient after 4920 hairs were transplanted in two sessions. (C) Schematic demonstrating the breakdown of the donor strip and the subsequent placement of those hair grafts.

Case Study 2
This patient is a 42-year-old white man who has diffuse Type VI thinning, thin, straight, and light brown hair (Figure 12A). He requests thickening of the frontal aspect of his scalp and wants to comb his hair straight back. After harvesting a 1.0 x 20.0 cm fusiform strip the following characteristics were present:

• 83181 total hairs (159 hairs/cm²);
• 1542 total follicular units (77 follicular units/cm²);
• 8516 one-haired follicular units (33% of all follicular units);
• 8509 two-haired follicular units (33% of all follicular units);
• 8421 three-haired follicular units (27% of all follicular units); and
• 96 four-haired follicular units (7% of all follicular units).

The strip was divided into the following grafts:

• 146 one-haired follicular units;
• 209 two-haired follicular units;
• 351 three-haired follicular units;
• 96 four-haired follicular units;
• 300 three haired follicular groupings; and
• 70 four-haired follicular groupings.

Figure 12C demonstrates the placement of the aforementioned grafts in the density distributions previously discussed in this article. As previously alluded to, it is the author's opinion that it is essential in this type of patient to make the incisions in between the preexisting follicular units with the help of 3.5 x expanded loupes. The following is the breakdown of the how the grafts were distributed:

• 100 one-haired follicular units for development of the frontal edge of the hairline;
• 46 one-haired follicular units for blending the posterior hairline abutting the vertex;
• 209 two-haired follicular units for gradation of the hairline;
• 351 three-haired follicular units at the lateral frontal areas;
• 96 four-haired follicular units centrally at the frontal area;
• 300 three-haired follicular groupings at the at the lateral posterior regions; and
• 70 four-haired follicular groupings at the central aspect of the posterior region.

After two sessions utilizing the above approach the results are shown in Figure 12B. One can notice that even though only 6215 hairs were utilized in this patient a fairly good result has been accomplished because the densely packed follicular units at the anterior aspect creates the illusion that the patient has more hair than he really has.

Figure 12. (A) The patient is a 42-year-old man with Type VI thinning. preoperatively. (B) Same patient after 6215 hairs were transplanted in two sessions. (C) Schematic demonstrating the breakdown of the donor strip and the placement of those hair grafts.

Case Study 3
The patient is a 39-year-old white man with Type VI diffuse thinning who has coarse, dark brown, wavy hair (Figure 13A). His goal is to wear his hair straight back. After harvesting a 1.0 x 20.0-cm fusiform strip the following characteristics were present:

• 2367 total hairs (118 hairs/cm²);
• 1336 total follicular units (67 follicular units/cm²);
• 528 one-haired follicular units (40% of all follicular units);
• 585 two-haired follicular units (24% of all follicular units); and
• 223 three-haired follicular units (16% of all follicular units).

The strip was divided into the following grafts:

• 198 one-haired follicular units;
• 255 two-haired follicular units;
• 223 three-haired follicular units; and
• 330 three-haired follicular groupings.

Figure 13C demonstrates the placement of the aforementioned grafts in the density distributions previously discussed in this article. The following is a breakdown of how the grafts were distributed:

• 90 one-haired follicular units to the anterior aspect of the frontal hairline;
• 84 one-haired follicular units to the posterolateral aspects of the grafted area;
• 24 one-haired follicular units at the posterior hairline abutting the vertex;
• 255 two-haired follicular units immediately behind the frontal one-haired follicular units for gradation;
• 223 three-haired follicular units directly behind the two-haired follicular units; and
• 330 three-follicular groupings centrally directly posterior to the three-haired follicular units.

Figure 13B demonstrates the results after two sessions utilizing the aforementioned techniques. Even though only 4621 hairs were utilized in this case, the results are significant because of the strategic placement of the grafts in congruence to the hairstyle desired.

Figure 13. (A) The patient is a 39-year-old man with Type VI thinning, preoperatively. (B) Same patient after 4621 hairs were transplanted in two sessions. (C) Schematic demonstrating the breakdown of the donor strip and the placement of those hair grafts.

 Discussion
As follicular unit transplantation has evolved over the years, there has also evolved a philosophy that surgeons should use nothing larger than a follicular unit during the hair transplantation process. This philosophy has developed because of the fact that the actual proportion of non-hair-bearing skin in a typical fusiform donor strip is probably on the order of 50%, which in turn causes a fair amount of non-hair-bearing skin to be included into any graft larger than a follicular unit. It has been stated that these larger hair grafts will negatively affect the aesthetics of the surgical hair restoration owing to the fact that a good fit into atrophic balding scalp will be difficult to achieve. This poor fit can cause tufting, dimpling, pigmentary changes; depression; or elevation of the grafts. It has also been stated that because larger incisions are needed for larger grafts there can be significant damage to the microvasculature of the recipient scalp.
 
The author does, in fact, agree with these basic premises, but disagrees when hair grafts have two or fewer closely juxtaposed follicular units. This article describes a technique that never uses more than four hairs per graft (which is in congruent with nature) and never uses more than two follicular units per follicular grouping. Additionally, the only follicular units that are incorporated into one graft are those that are naturally closer together than the others within the donor strip.
 
Using the approach discussed in this article accomplishes five goals: (1) it allows the procedure to be performed at a much quicker pace owing to less dissecting and less placing; (2) it lessens iatrogenic injury to follicular units during the dissection phase; (3) it lessens iatrogenic injury during the placement phase because follicular groupings are a little less fragile than a follicular unit because more tissue is present; (4) it allows the grafts to be left out of the body for a shorter period of time which increases survival rates; and (5) it is less costly to the patient.
 
The fact that the procedure time is lessened cannot be overlooked as being unimportant. Limmer17 demonstrated, in an excellent study, that the longer grafts are left out of the body, the lower the survival rate (Table 1). In his study, there was 95% survival at 2 hr, 90% survival at 4 hr, 86% survival at 6 hr, 88% survival at 8 hr, 79% survival at 24 hr, and 54% survival at 48 hr.

Table 1. Hair Survival in Micrografts (Two Cases)
Number of Hours between Harvest and Transplantation	Number of Transplanted Hairs	Number of Hairs Surviving at 5 1/2 Months after Transplantation (%)
2	257	244 (95)
4	200	180 (90)
6	200	173 (86)
8	227	219 (88)
24	200	158 (79)
48	200	109 (54)

Besides survival rates being affected by increased time, one must also account for the fatigue factor as procedures get longer and longer. As surgical assistants get tired, the dissection will not be as accurate and there will be more iatrogenic damage to the hair grafts.
 
In regard to the argument that larger incisions cause more trauma to the scalp, the author believes that the technique described in this article does not increase trauma in any way. In fact, it can easily be demonstrated that there is actually less trauma when the longer 2.0-mm-long blade incisions are made for three and four-haired follicular groupings. For example, when two-, three-, or four-haired follicular units are utilized, the author usually makes 20 1.5-mm incisions in a square centimeter. Therefore, when the total length of the incisions is cumulated a length of 3.0 cm results (1.5 mm x 20 = 30 mm = 3 cm). Conversely, when the author uses three- or four-haired follicular groupings, he makes eight incisions of 2.0 mm in length in a square centimeter. Thus a total of 16 mm or 1.6 cm of incisional length is created in a square centimeter (2.0mm x 8 = 16mm = 1.6cm).
 
In this article, 5362 hairs were transplanted in Case 1 during two sessions, 6215 hairs in Case 2 with two sessions, and 4621 hairs in Case 3 with two sessions. With each of these cases, all three still have plenty of hair for further surgical hair restoration. To illustrate this numerically, it is well known that the average head contains 100,000 hairs at peak density, with approximately 25% being at the donor dominant area. Therefore, 25,000 hairs are permanent with one-half of these (12,500) being available for transplanting. In Case 1 we have used approximately 43% of available donor hair; in Case 2 approximately, 50% of the available donor hair; and in Case 3, approximately 37%. Although all three of these patients had hair densities that were less than average, these three examples give approximate illustrations as to how each of these individuals will have adequate donor hair in the future to deal with senile alopecia or future male follicular pattern baldness. This was made possible primarily by the strategic placing of certain hair grafts in vital areas based on the patient's desired hairstyle.
 
Summary
A philosophy has developed in surgical hair restoration that nothing larger than a follicular unit should be utilized in the hair tranplantation process. It is the message of this article that when two follicular units are extremely close together, it is better to make a three- or four-haired follicular grouping instead of trying to divide the two follicular units. By utilizing this approach the surgeon will accomplish five things: (1) the procedure will be performed more quickly owing to less dissection; (2) there will be less iatrogenic damage to the hair follicles during the dissection phase; (3) there will be less iatrogenic to the hair follicles during the placement phase owing to the fact that follicular groupings are less fragile than follicular units; (4) it allows the hair grafts to be left out of the body for a shorter period of time which increases survival rates; and (5) the procedure will be less costly to the patient.
 
References
1. Okuda S. The study of clinical experiments of hair transplantation. Jpn J Dermatolurol 1939;46:135.
2. Tamura H. Hair grafting procedure. Jpn J Dermatol Venereol 1943;52.
3. Fujita K. Hair transplantation in Japan. In Kobor I, Montagna,  eds. Biology and Disease of the Hair. Baltimore University Park Press, 1976:519-27.
 4. Orentreich N. Autografts in alopecia and other selected dermatologic conditions. Ann N Y Acad Sci 1959;83:463.
5. S. Ayers S III Hair transplantation for male pattern baldness, aesthetic considerations and current status. Head Neck Surg 1985;7:272-85.
6. Brandy DA. Conventional grafting combined with follicular groupings: a new approach. J Dermatol Surg Oncol 1987;13:60-3.
7. Orentreich N, Orentreich DS. Hair transplantation. J Dermatol Surg Oncol 1985;11:319-24.
8. Nordstrom R. Micrografts for the improvement of the frontal hairline after hair transplantation. Aesth Plast Surg 1985;5:97.
9. Bradshaw W. Quartergrafts,: a technique for follicular groupings. In: Unger WP, Nordstrom R, Rolf EA, eds. Hair Transplantation, 2nd ed. New York: Dekker, 1988.
10. Headington JT. Transverse microscopic anatomy of the human scalp. Arch Dermatol 1984;120:449-56.
11. Limmer BL. Elliptical donor stereoscopically assisted micrografting as an approach co further refinement in hair transplantation. Dermatol Surg 1994;20:789-93.
12. Seager D. Binocular stereoscopic dissecting microscope: should we use them? Hair Transplant Forum Int 1996;6:2-5.
13. Rassman WR, Bernstein RM, Szaniawski W, et al. Follicular transplantation. Int J Aesth Restor Surg 1995;3:119-32.
14. Rassman WR, Bernstein RM. The aesthetics of follicular transplantation. Dermatol Surg 1997;23:785-99.
15. Bernstein RM, Rassman WR. The logic of follicular unit transplantation. Dermatol Mawl Clin 1999;17:277.
16. Brandy DA. A technique for hair-grafting in between existing follicles in patients with early pattern baldness. Dermatol Surg 2000; 26:801-5.
17. Limmer BL. Micrograft survival. In: Stough DB, Haber RS, eds. Hair Replacement: Surgical and Medical. St. Louis: Mosby, 1996:147-9.
 
Commentary
The concept of "cherry-picking" grafts that contain two follicular units, by looking for follicular units in the donor tissue, that are closer together than usual, has been advocated at meetings for the past several years-most notably and frequently by William Reed and Michael Beehner. The wisdom of doing this as well as very useful details of technique is elaborated on in this article by Brandy. A few comments are warranted. (1) Although follicular units are an average of 1 mm apart, they can be as far as 1.5 mm apart or close enough together (0.2 mm) that they look almost like a single follicular unit.¹ As indicated above, those that are closer together than average can be used as double-follicular-unit grafts that are placed into the small slits described by Brandy. It is this type of graft that he refers to as "minigrafts" (rather than double follicular units) in his article. The follicular units that are close enough together to look like a single follicular unit, on the other hand, provide us with an additional (not necessarily alternate) advantage: they can be left together as a type of double follicular unit that is small enough to fit into a 16- to 18-gauge needle hole. Such grafts were first described by Seager² and were named "follicular families." More recently, Tykocinski¹ has elaborated on their use and suggested an alternate name; "follicular grouping grafts." The advantage of utilizing these follicular families (or follicular grouping grafts) is that it thereby increases the number of grafts containing three, four, and five hairs that can be transplanted at least as densely as Brandy typically transplants his follicular units-20 to 25/cm². Thus, the density of hair in specifically chosen areas can be even greater than Brandy describes here for his follicular unit transplanting. Since Seager first described follicular families, an increasing number of practitioners (including this reviewer) have routinely been looking for these follicular families and incorporating them into our grafting when the hair is fine textured and/or there is little hair/skin color contrast. (They might otherwise look too "pluggy.") Twenty-five of these grafts per square centimeter, for example, each of which contains four hairs, could result in the growth of 100 hairs/cm² after a single session. Shapiro,³ Harris,⁴ and others have also described "follicular pairing" in which two individually prepared follicular units are placed into a single needle hole to achieve similar goals. This concept is described, in detail, in the fourth edition of Hair Transplantation. (2) Although Brandy refers to 25 follicular units/cm² as being "densely packed," other authors define "dense packing" as 40, 50, or even 70 follicular units/ cm². The reader therefore is urged to be sure of the definition of "dense packing" by whomever uses this terminology in other presentations. This is important because there is likely to be a significant difference in how many hairs survive in areas treated with 70 follicular units/cm² when compared with how many survive in areas treated with 20 or 25 follicular units/cm2-at least in most offices. In general, most practitioners are not satisfied with the hair yield that they achieve with densities greater than 30 follicular units/cm² and a majority probably do not use anything over 25 follicular units/cm². Mayer, for example, showed substantial and decreasing hair survival rates in grafts he transplanted at 10, 20, 30, and 40 follicular units/ cm² (Mayer MA, "Follicular Regeneration," presented at the annual meeting of the ISHRS, Washington, DC, September 17, 1998). (3) Although Brandy uses his three- and four-haired follicular units in a zone that runs posterior to the entire length of the hairline zone, other authors use most of these grafts in adifferent distribution-most commonly, an egg-shaped area in the midline, posterior to the hairline zone.³ They will use some three- and four-haired grafts in the same location as Brandy does, but they save the majority of them for this egg-shaped area, believing that it more effectively creates an illusion of density in the frontal area. I am among those who prefer this latter type of graft distribution. (4) Brandy quotes Limmer's hair survival studies, which demonstrated 88% hair survival of grafts transplanted after 8 hr, but 95% at 2 hr.⁵ It should be noted that these hair count studies were performed 6 months after transplantation and we are now aware of hair growth occurring as late as 18 months after surgery (Martinick J, "The Results at 18 Months of the Longitudinal Clinical Research into the Importance of Transplanting Intact Follicular Units vs. Follicular Units That Have Been Traumatized Using a Variety of Methods Including Transection at the 'Bulge,'" presented at the 8th annual meeting of ISHRS, Hawaii, December 2000). Limmer has commented that the results of his studies may have been substantially different if he had waited for 12 months to perform his hair counts. Incidentally, the same study that showed hairs growing as late as 18 months after surgery, also showed that follicular units that were sectioned longitudinally, for example, separating one hair of a three-hair follicular unit away from the unit, grew as well as those hairs in intact follicular units. (5) The difference in density that is possible, after a single session with the use of double follicular units, as opposed to follicular unit transplanting, can be minimized by adding single follicular units between the double follicular units and, in fact, the density of follicular units per square centimeter with a combination of follicular units and double follicular units can be equivalent to that achieved with 20 to 25 follicular units/cm² planted as individual follicular units.⁶ The advantage of using the combination of follicular units and double follicular units instead of follicular unit transplanting is that the number of follicular units can be doubled or tripled with a second or third session of the former but not the latter. There are other advantages to combining follicular units and double follicular units (or even triple follicular units) if one wants to ultimately produce an appearance of even more density than has been created with follicular unit transplanting to date. Unfortunately, space does not allow for elaboration of that here, but it is also covered in detail in the fourth edition of Hair Transplantation;⁶ once again, the photos Brandy shows in this article do not adequately demonstrate this potential. On the other hand, transplanting with single follicular units will consistently produce a more even coverage of hair, especially after a single session in an alopecic area. (6) I, among others, doubt the estimate that 25 % of hair in the donor area and therefore only 12,500 hairs are available for transplanting in the average patient. It is my view that both numbers are substantially incorrect and unnecessarily limit what can be accomplished in the recipient area. This is especially true if one excises old donor area scars as a component of every subsequent session, as Brandy has recommended elsewhere. In my experience, if only a single scar is created, regardless of the number of sessions carried out, one can often harvest far more than 25% of the hair in the donor area without any cosmetically significant thinning of hair in that area.
 
Walter Unger, MD
Toronto, Ontario, Canada
 
References
1. Tykocinski A. A one-year study of using "follicular grouping grafts” in specific areas.... Hair Transplant Forum Int 2003;13365:368-9.
2. Seager D. Dense hair transplant from sparse donor area-introducing the "follicular family unit". Hair Transplant Forum Int 1998;8:21-2.
3. Shapiro R. Follicular unit transplantation (FUT) alone or follicular units with multi-FU grafts-why, when, and how. In: Unger WP, Shapiro R, eds. Hair Transplantation, 4th ed. New York, Dekker, 2003:in press.
4. Harris J. Recombinant follicular units-concept formalization. In: Unger WP, Shapiro R, eds. Hair Transplantation, 4th ed. New York: Dekker, 2003:in press.
5. Limmer BL. Micrograft survival. In: Stough DB, Haber RS, eds. Hair Replacement: Surgical and Medical. St. Louis: Mosby, 1996:147-9.
6. Unger WP. Why "mixed" grafting: follicular units and multi-FU grafts. In: Unger WP, Shapiro R, eds. Hair Transplantation, 4th eds. New York: Dekker, 2003:in press.

to the top